9727 - Laserfiche WebLink

Home Browse Search

200 wild adult ra- lakeside backwater Ig season produced stripping gametes, 200,000 embryos )duced only 17 ju- inckley et al. 1991; .ffort began in 1994 honed to gather nat- ets and bays around al. 1993). To date tan 440,000 larvae. protected lakeside facilities that were td after achieving a les were measured, i transponder; Bio- id repatriated to the tpatriation program 1, to date, with an- ,ild razorback suck- the early 1970s un- inckley at Arizona ng effort has pro- stimates that have )ng with the status nckley et al. 2003), A the group's goal 50,000 razorback wever, the program the genetic charac- i have been passed ite larvae (Dowling present population th rate estimates of r repatriation pro- 2; assess the suc- ;ram from a demo- recommendations. -patriated fish were nmingled with wild tg grounds in the censuses of adults .th peak razorback ,hen fish were ag- and along shallow nckley 1983). Less erred at nonspawn- ts (20-100 m long r mesh, and 25.4- t-mounted electro- REPATRIATION TO CONSERVE IMPERILED FISH fishers were the primary sampling devices, al- though gill nets, hoop nets, and seining with tram- mel nets also were used. Overnight net sets and electrofishing were in water 2-6 m deep, and nets were cleared of fishes at 4-6-h intervals. Razor- back suckers were placed into onboard live tanks or into floating enclosures until assessment. With minimal handling before release, fish were mea- sured (TL), examined for sex and general health, and electronically scanned for PIT tags to deter- mine census data (capture [first-time capture] or second [or third, etc.] recapture). Sex categories were defined as juvenile (a young fish that did not exhibit external secondary characters), female, male, and unknown (adult but sex could not be determined reliably). All repatriates were released near their site of capture. Although sampling focused on a small portion of the entire lake, we are confident that our samples are representative of the Lake Mohave razorback sucker population as a whole. Extensive sampling lakewide indicated that in spring few adult razor- back suckers are found in Lake Mohave except on spawning grounds (Bozek et al. 1984; Minckley 1983; Minckley et al. 1991; unpublished data). This result has been confirmed by telemetry stud- ies (Mueller et al. 2000), which also demonstrate that fish move freely among spawning areas. Fur- thermore, low-level aerial surveys conducted as frequently as biweekly during the reproductive pe- riod for at least the past decade (i.e., 1990s) also confirm-that most fish are found in association with spawning aggregations, which comprise the rela- tively few places where netting is done. Annual single-census population estimates (N`) were determined for repatriated razorback suckers by applying paired March census data to a mod- ified Peterson method formula (i.e., Chapman modification; Seber 1973), namely, N*=(M+1)(C+1)_ 1 R+1 We modified M, C, and R definitions in Ricker (1975) because all repatriates were tagged before stocking and, thus, were marked when first en- countered; therefore, M = number of repatriated fish captured the first year, C = number of repa- triated fish captured the following year, and R = total number of repatriated fish captured in both the first year and following year. Population esti- mates were reported for the marking (first) year because only natural mortality and emigration losses were assumed to occur (Seber 1973). Fish 549 captured and recaptured within the same sampling period were eliminated from analysis. Confidence intervals were derived using Poisson approxima- tion tables (Seber 1973) using R as the entering variable. Values from the table were multiplied by M X R to obtain the appropriate confidence inter- val. Survival rates were calculated by dividing the annual single-census population estimates by the cumulative number of repatriates released. A study model should meet three assumptions when applying the modified Peterson estimate (Pollock et al. 1990): (1) the population is closed to both additions and deletions, (2) no tags are lost or omitted, and (3) equal catchability of all indi- viduals. In our study, natural mortality of repatri- ates was occurring, but this apparent violation of the closure assumption did not bias estimates be- cause only one process (deletion) was in effect. Mortality was not generally associated with PIT tag usage, and no natural recruitment has been detected in recent decades in Lake Mohave (Minckley 1983; Minckley et al. 1991) or during the period of this study. Tag loss was assumed negligible or zero because PIT tags were consid- ered permanent (Prentice et al. 1990). However, equal catchability could not be assumed because sampling was restricted to spawning areas and per- formed primarily with trammel nets. Juvenile ra- zorback suckers typically did not occupy the same habitat as spawning adults (Mueller et al. 2000), and electrofishing and trammel nets tend to be size- selective for larger individuals, thus, resulting in unequal catchability between larger adults and smaller juvenile fish. Similarly, repatriated fish captured in the marking year are probably less available for recapture in the census year if they still are immature, and conversely, juveniles that mature before the census year have an equal op- portunity for capture in the census year. Both of these possibilities act to increase the proportion of captured to recaptured fish in the census year, which would inflate the population estimate. How- ever, survival estimates will underestimate true survivorship because a majority of fish released in the 1-3 years preceding the estimate will not be available for capture during the interval. Survival as a function of size at release has long been suspected as a major contributor to past fail- ures in the razorback sucker repatriation program (Marsh and Brooks 1989; Minckley et al. 1991). We used log-linear analysis as described by Sokal and Rohlf (1995) to determine the effect of size at release on overall survival and the consistency of this effect among years. Size at release was