Estimating Fish Populations by Removal Methods with Minnow Traps in Southeast Alaska Streams

-Passive capture methods, such as minnow traps, are commonly used to capture fish for mark-recapture population estimates; however, they have not been used for removal methods. Minnow traps set for 90-min periods during three or four sequential capture occasions during the summer of 1996 were used to capture coho salmon Oncorhynchus kisutch fry and parr, Dolly Varden Salvelinus malma, cutthroat trout O. clarki, and juvenile steelhead O. mykiss to estimate population size with the Zippin or generalized removal method. More than 45% of the total catch was obtained during the first capture occasion, and in most cases, the catch during the fourth occasion was less than 15% of the total catch. In most pools, the probability of capture was greater than 0.4 but was lower for coho salmon fry than for coho salmon parr and other species. Mean population estimates for coho salmon parr made with concurrent mark-recapture and removal methods differed significantly in small streams. Estimates from mark-recapture and removal methods were not significantly different for coho salmon fry and Dolly Varden, but mark-recapture estimates were higher than removal estimates in most cases. My results show that removal estimates can be obtained with minnow traps if sampling procedures conform to the assumptions required for the method. Obtaining precise and accurate estimates of fish abundance in streams continues to challenge fishery biologists, despite the development of sophisticated mathematical models. Commonly used methods include mark-recapture experiments (Ricker 1975; Zubik and Fraley 1988) and removal estimates (Moran 1951; Zippin 1958; White et al. 1982). Though snorkel surveys are also used to estimate fish abundance (Northcote and Wilke 1963; Schill and Griffith 1984; Thurow 1994), they require a separate estimate of the population to calibrate the counts (Hankin 1986). Mathematical models for both mark-recapture and removal estimates are well-tested, but present substantial logistical challenges to meet the assumptions. Mark-recapture estimates are commonly used in southeast Alaska and elsewhere to estimate Populations of juvenile salmonids, most commonly coho salmon Oncoryhnchus kisutch and Dolly Varden Salvelinus malma, in small (<4-m-wide) secondto third-order streams (Elliott and Hubartt 1978; Dolloff 1983; Bryant 1984; Young et al. 1999). Sample reaches in streams wider than 4 m and with higher water flows are difficult to isolate, and mark-recapture methods are not reliable because of movement between sample periods. High flows, common in southeast Alaska, also affect movement and catchability between sample periods. Removal methods or snorkel surveys are often used in these streams, yet even these methods are limited. Low conductivity and patches of complex habitat with large woody debris make the removal method of electrofishing impractical. Snorkel surveys also are impractical because of complex habitat and poor visibility in the dark waters of many southeast Alaska streams. Removal methods have several advantages over mark-recapture methods to estimate fish numbers. Fish are captured only once, which eliminates bias due to behavioral responses to a trap. Fish do not need to be marked, which removes assumptions that all marks are identified and that negligible mortality occurs due to marking. The stream section can be sampled in 1 d, which substantially reduces the probability of movement by fish into and out of the sample area in cases in which the stream section cannot be isolated for the duration of the mark-recapture sequence. In addition, a 1 d sampling effort simplifies logistics for those locations that are difficult to reach and eliminates any differences in sampling efficiency due to changes in flow regimes (i.e., high-water events that occur after marking and before or during recapture). Passive capture methods are commonly used for mark-recapture experiments but are seldom used for removal estimates. Minnow traps baited with salmon eggs are an effective method for capturing juvenile salmonids and have been used in numer* E-mail: mdbryant@ fs.fed.us Received April 12, 1999; accepted March 30, 2000 North American Journal of Fisheries Management 20:923-930, 2000 © Copyright by the American Fisheries Society 2000