7308 - Laserfiche WebLink

Home Browse Search

~. I USE OF RADIOTELEMETRY 179 aken to minimize dis- er depth was recorded ith a wading rod, and red 0.6 distance below lrsh-McBimeycurrent obtained by direct ob- with a wading rod. 60 60 GREEN - COll. 30 n.128 GREEN -RADIO 30 n.1405 to reduce the bias in ng from every identi- studied were divided lJS sections of fish hab- ~eomorphology. With- tations were selected rtbers. Habitats within d using electrofishing, wire traps, depending :ear type. The habitat point of capture for were recorded, and 'ere measured as pre- .... z w 0 U ~ 60 ~ All - COlL. 30 n:241 60 All - RADIO 30 n.2329 o Shore eddy backwat.r pool other ,HABITAT hSCUSSION reference was record- 22 radiotelemetered ~ in total length from mm). Because of the iQuawfish (Tyus et al. It information from ~ river. One fish was .uchesne, and White e fish was monitored vers and one in the 'ig. I). Of the 22 fish, freen River, 7 in the (ampa River. In ad- was recorded at the trado squawfish 405- Colorado squawfish habitats but mostly Figure 2. Habitat types recorded for collected (COLL) and radiotelemetered (RADIO) Col- orado squawfish from the Green River in 1980 and 1981, and the Green, White, and Yampa rivers (ALL) in 1981. (Miller et al. 1982; Miller et al. 1982) with 2,329 radiotelemetry observations from all three rivers (Fig. 2) produced this same relationship. The dif- ference between the two methods apparently is due to bias in habitat types recorded from col- lected fish because most of the collections were made by electro fishing (Hynes 1970). When Col- orado squawfish in shoreline habitats are ex- posed to an electric current, they frequently were observed attempting to escape into deeper water and moved into runs where they were captured. For this reason, it appears that the radiotelem- etry data were more representative than that tak- en from fish captured by electro fishing. Colorado squawfish were found mainly over sand substrate; however, they were also found over silt, rubble, boulder, and gravel depending on the time of year (Fig. 3). Sand substrates were difficult to sample adequately with conventional gear because nets are hard to set in mid-channel areas, and fish can escape more easily from the fringes of the electric current (Hynes 1970). Also, d at the point of col- r for 128 Colorado I (Tyus et al. 1982) observations of ra- ~sultant histograms more collected fish from shoreline hab- ltelemetry observa- f data from 24 I fish ~, and Yampa rivers COll ",10B .... 0 z w U a:: w ~ RADIO ",908 SAND SILT BOULDER RUBBLE GRAVEL SUBSTRATE Figure 3. Percentage substrate types recorded for collected (COLL) and radio telemetered (RADIO) Colorado squawfish. - radiotelemetry indicated the fish, at times, se- lected drop-offs next to sand bars in the mid- river channel. In this situation, radiotelemetry was the only method that could be relied upon to produce accurate habitat preference infor- mation. Unfortunately, both methods were biased because investigators could not determine the precise substrate composition of deep-water hab- itats. Water depths and velocities recorded at the observed locations of radiotelemetered Colorado squawfish (Fig. 4) indicated 13 fish in the Green River selected an average water depth of 1.4 m and an average velocity of 0.2 m/second, 7 fish in the Yampa River selected 0.94 m and 0.12 m/second, and 6 fish in the White River pre- ferred 0.72 m and 0.49 m/second. Analysis of variance indicated no significant (E = 0.01) dif- ference in the depths recorded from the Green River between 1980 and 1981. There was a sig- nificant difference (P = 0.01) between depths re- corded from the Green River and its two trib- utaries (White and Yampa rivers) but depths recorded between the White and Yampa rivers were not significantly different (P = 0.0 I). These results are supported by similar tests of data ob- tained by conventional fish collections.