7308 - Laserfiche WebLink

Home Browse Search

- ~ 178 TYUS, BURDICK, AND MCADA f ! N 1 1--1 L.. SCALE IN 'ULOIIIETEAS 50 100 150 200 250 WYOMING 50 ,.. SCALE IN MILES UTAH COLORADO I I --------------t------------ I i ARIZONA NEW MEXICO Figure 1. Location of study rivers in the Green River Basin. and one period was picked at random. Beginning with the selected 8-hour period, the selected fish was observed for three 24-hour periods and its location was recorded every 15 minutes. After the fish had been observed for three 8-hour pe- riods, the fish with the next tag number was se- lected for study. The location of each radiotelemetered fish was carefully fixed. If it remained in one location for 30 minutes, it was assumed that this was the preferred habitat and microhabitat information was recorded, including the general habitat and substrate type, water depth, and velocity. Gener- al habitat types included: Shorelines-shallow, low-velocity waters next to shore. Eddies-deep shoreline whirlpools with up- stream velocity. Runs-channels with swift laminar flow. Backwaters-semi-isolated water bodies with no measurable velocity. Pools-deep, quiet portions of the stream. Water depth, velocity, and substrate measure- ments were taken only when the fish moved to another location or at the end of the study period. These precautions were taken to minimize dis- turbance to the fish. Water depth was recorded by direct measurement with a wading rod, and water velocity was measured 0.6 distance below the water surface with a Marsh- McBimey current meter. Substrate type was obtained by direct ob- servation and by probing with a wading rod. Conventional Collections An attempt was made to reduce the bias in fish collections by collecting from every identi- fiable habitat type. Rivers studied were divided into relatively homogeneous sections offish hab- itat based on general river geomorphology. With- in each section, 0.8-km stations were selected from a table of random numbers. Habitats within these stations were sampled using electrofishing, trammel nets, seines, and wire traps, depending on the suitability of each gear type. The habitat and substrate types at the point of capture for each Colorado squawfish were recorded, and water depth and velocity were measured as pre- viously described. RESULTS AND DISCUSSION Information on habitat preference was record- ed from observations of 22 radiotelemetered Colorado squawfish ranging in total length from 385 to 707 mm (.f = 535 mm). Because of the high mobility of Colorado squawfish (Tyus et al. 1982), we recorded habitat information from three fish in more than one river. One fish was monitored in the Green, Duchesne, and White rivers in 1980; in 1981, one fish was monitored in the Green and White rivers and one in the Green and Yampa rivers (Fig. I). Of the 22 fish, 13 were monitored in the Green River, 7 in the White River and 6 in the Yampa River. In ad- dition, habitat information was recorded at the point of capture for 241 Colorado squawfish 405- 707 mm long (.f = 50 I mm). Colorado squawfish were found in a variety of habitats but mostly along shorelines (Fig. 2). The habitat types recorded at the point of col- lection in the Green River for 128 Colorado squawfish in 1980 and 1981 (Tyus et al. 1982) were compared with 1,405 observations of ra- diotelemetered fish. The resultant histograms were quite similar, except more collected fish were taken in runs and less from shoreline hab- itats than recorded for radiotelemetry observa- tions. Further comparisons of data from 241 fish collected in the Green, White, and Yampa rivers 60 30 60 30 I- Z W 0 U ~ 60 Q. 30 Shore .dely baCl HABITAT Figure 2. Habitat types reo (COLL) and radiotelemetel orado squarish from the G and 1981, and the Green, , rivers (ALL) in 1981. (Miller et al. 1982; Miller et a radiotelemetry observations f (Fig. 2) produced this same reI ference between the two met due to bias in habitat types I lected fish because most of tl made by electrofishing (Hyne1 orado squawfish in shorelin posed to an electric current, tl observed attempting to escap and moved into runs where t For this reason, it appears n etry data were more represent en from fish captured by ele( Colorado squawfish were j sand substrate; however, the over silt, rubble, boulder, anl on the time of year (Fig. 3). S. difficult to sample adequatel) gear because nets are hard to areas, and fish can escape ill< fringes of the electric current (