9726 - Laserfiche WebLink

Home Browse Search

1252 MARSH ET AL. volunteers. Sampling sites were identified through observation of adult razorback suckers consis- tently visiting specific areas of shoreline to stage and spawn in springtime. Primary sampling sites and their approximate river kilometer (RK) up- stream of Davis Dam on the Arizona side of Lake Mohave included Owl Cove (RK 45), Yuma and Gold coves (RK 39), Arizona Bay (RK 38), Carp Cove (RK 32), and Cottonwood Cove East (RK 32), while sites on the Nevada side included the area between Cottonwood Cove West and Tequila Cove (RK 35), Hog Farm Cove (RK 31), Six Mile Cove (RK 30), the area between Nine Mile and Half-way coves (RK 27), and Nine Mile Cove (RK 26). Annual spring (March) sampling occurred in the above-mentioned coves and inlets along shallow shorelines, reefs, and spits, with variously sized trammel nets (20-100 m long X 3 m deep, 2.5- 3.8-cm inner mesh, and 25.4-30.5-cm outer mesh) and boat-mounted electrofishers. Gill nets and hoop nets were used infrequently. Overnight net sets were in water 2-6 m deep, with nets cleared of fishes at 4-6-h intervals. Electrofishing was typ- ically performed at night over similar depths. Ra- zorback suckers were placed into on-board live tanks or into floating live enclosures until assess- ment. With minimal handling, fish were measured to the nearest millimeter (total length [TL]) and examined for sex, parasites, disease, deformity, in- jury, and general health. Sex categories were de- fined as juvenile (a young fish that has not attained sexual maturity and does not exhibit external sec- ondary characters that allow reliable sex deter- mination), male, female, and unknown (an adult fish whose sex could not be reliably determined). All razorback suckers were electronically scanned for passive integrated transponder (PIT) tags (Biomark, Inc., Meridian, Idaho). Most fish cap- tured unmarked were injected with a PIT tag, but occasionally fish were released unmarked, al- though their TL, sex, and physical data were re- corded. Approximately 100 fish were removed from the population during the last decade for their reproductive services at national fish hatcheries or for their participation in telemetry and other stud- ies (e.g., Minckley et al. 1991; Mueller et al. 2000; Marsh, unpublished report). Annual single-census population estimates were determined for razorback suckers by using an ad- justed Peterson-method formula (i.e., N*, the single- census Chapman modification; Ricker 1975): N* _ ([M + 11 X [C + 11)1(R + 1), where N* is the annual single-census population estimate for the marking or previous year, M is the number of fish marked during the marking year, C is the combined number of marked and unmarked fish captured in the sampling or current year, and R is the number of marked fish captured in the sampling year that were originally marked in the marking year. C and R do not include multiple captures of the same marked fish during the same sampling year. Mortality was assumed to be greater than zero and recruitment was assumed to be zero, so N* was calculated for the marking year rather than the sampling year (Seber 1973). Confidence limits (CLs) of N* were based on a Poisson fre- quency distribution with R as the entering variable x in the distribution table; if R was greater than 50, Pearson's formula was used (Ricker 1975): x + 1.92 ± 1.960. (x + 1.0)0.5 Linear regression was used to determine wheth- er M and R were significantly related to time. All statistical analyses were based on a significance level of 0.05. Our study followed three basic assumptions, generalized by Pollock et al. (1990) for the ad- justed Peterson estimate: (1) the population is closed, with no additions or deletions; (2) all an- imals have a similar likelihood of capture in each sampling year; and (3) marks are not lost or over- looked. The first assumption is less stringent if mortality is random and does not differ between marked and unmarked fish. Fish in our study were subjected to two processes, capture and handling. Razorback suckers that were found dead in nets or that were observed dead after handling were a rar- ity; however, results indicate occasional substan- tial annual losses, and a zero-mortality assumption for the adjusted Peterson method was clearly vi- olated. The second assumption requires that the capture probability for individuals is homogenous throughout the population and that fish do not ex- hibit a "trap response." Homogeneity was as- sumed for our study population for several reasons, the first of which was based on the similar age and size of individuals in the population (McCarthy and Minckley 1987). Also, sampling was focused on spawning sites, and all adults were assumed to participate in spawning behavior. Further, razor- back suckers have been shown in the past to dis- perse great distances after release (Mueller et al. 2000) as well as to randomly mix into the popu- lation. Finally, the possibility of a trap response is rejected because trammel nets are not baited;