This file was created by scanning the printed publication. Ecology of Freshwater Fish 1997: 6: 1-7 Errors identified by the software have been corrected; Copyrixht 0 MunksRaurd 1997 Printed in Denmark . All rights reserved however, some errors may remain. ECOLOGY OF FRESHWATER FISH 1SS.V 0906-6691 Habitat utilization and diel behavior of juvenile bull trout (Salvelinus conflwnw) at the onset of winter Thurow RF. Habitat utilization and diel behavior of juvenile bull trout R. F. Thurow (Sulvelinus confluentus) at the onset of winter. US. Forest Service, Intermountain Research Ecology of Freshwater Fish 1997: 6: 1-7.0 Munksgaard, 1997 Station, Boise, Idaho, USA Abstract - Underwater observations were used to describe habitat use and diel behavior of juvenile bull trout (Salvelinus confluentus) at the onset of winter (0.8”C) in a second-order central Idaho stream. All fish observed during day- time counts were concealed beneath “home stones” (mean dimensions 32.3 by 21.9 cm) in primarily pool and run habitats. Focal point depths and velocities averaged 57.2 cm and 4.7 cm/s (home stone removed). Fish maintained sta- tions over substrate with low percentages of fine (<2 mm, mean=5.2%)sedi- ment. Concealed bull trout were not evenly distributed across the stream chan- nel. Significantly more home stones were distributed near the midline of the channel, compared to the channel margins. Fish size was positively correlated Key words: habitat utilization; die1 behavior; bull with home stone size. At night, bull trout exhibited a diel behavioral shift, trout some fish moved out of daytime concealment cover into the water column. At Russell F. Thurow, U.S. Forest Service, night, bull trout were observed feeding and resting, primarily in pool and run Intermountain Research Station, 316 E. Myrtle, habitats. During both day and night, bull trout used riffles significantly less Boise, ID 83702, USA frequently than those habitats were available. Accepted for publication January 19,1996 Un resumen en espaiiol se incluye detras del texto principal de este articulo. introduction (Edmundson, Everest & Chapman 1968; Hillman, Griffith & Platts 1987; Cunjak 1988). Overwinter ecology of stream-dwelling salmonids Despite increased investigation of overwinter is perhaps the least understood aspect of their life habitat during the past decade (McMahon & Hart- history, and the need for winter investigations has man 1989), knowledge of the winter behavior and long been recognized (Hubbs & Trautman 1935). ecology of important native salmonids is incom- Many species of juvenile salmonids occupy differ- plete. In one of the few winter studies of bull trout ent habitats in winter than in summer (Hartman (Sulvelinus conjuentus), Jakober (1995) observed a 1965; Everest 1969; Bustard & Narver 1975; Rim- diel shift in behavior and cover utilization as water mer, Paim & Saunders 1983). At the onset of winter, temperatures declined. Other researchers have ob- stream-dwelling salmonids in the Intermountain served bull trout concealed during winter days and West typically adopt two overwintering strategies, resting on the substrate during winter nights (Bon- migration and concealment. As water temperatures neau 1994; Goetz 1994). decline below 10°C, fish may migrate from summer Bull trout are the only char native to the Inter- habitat into other portipns of watersheds (Bjornn mountain West. Some local populations are extinct, 1971). Distances fish move may be influenced by and many populations have declined throughout the the proximity of suitable overwintering habitat; range (Rieman & McIntyre (1995). Concern for the some fish migrate more than 100 km (Bjornn & persistence of bull trout culminated in petitions for Mallet 1964). After locating suitable overwinter review of the species status, and in June 1995, the habitat, juvenile salmonids typically select areas of U.S. Fish and Wildlife Service determined that list- low water velocity and enter concealment cover ing under the Endangered Species Act was war- 1 Thurow ranted but precluded. Data are necessary to describe veys were made between 1000 and 1600 h. When the winter ecology and critical habitats used by bull no fish were visible, I swam back to the start of the trout. This article describes the characteristics of habitat and zigzagged upstream while randomly se- overwinter habitat used by juvenile bull trout and lecting substrate and overturning it in search of fish. their die1 behavior in a second-order stream. I applied a similar sampling effort in each habitat and overturned more than 100 stones. If no fish were encountered beneath the randomly selected Study area substrate, I did not record any information and Bull trout were observed in Profile Creek, a tribu- moved upstream to another location. tary to the East Fork of the South Fork Salmon When I encountered a fish beneath the substrate, 1 River near Yellowpine, Idaho. Profile Creek flows marked its focal point with a weight and flag (Fausch through the central Idaho Batholith, an area of gra- &White 1981). I estimated fish size by approaching nitic bedrock characterized by steep slopes and fish, aligning their snout and tail with adjacent ob- highly erodible soils (Megahan, Platts & Kulesza jects and measuring the distance with a rule (Cunjak 1980). Peak stream discharges are caused by snow- & Power 1986a). Bull trout were classified into 50- melt, typically in May or June. August 1991 dis- mm length groups, excluding age-0 fish. Griffith charge was 0.71 m3/s. Base flows occur from Sep- (198 1) reported that trained underwater observers tember through January. Most annual precipitation were able to estimate fish size within 25 mm of the falls as snow and may exceed 1.5 m. Freezing tem- true length in most trials. An assistant on shore fol- peratures can occur every month. Winter air temper- lowed, called out habitat boundaries, and recorded atures occasionally drop below -5°C. data. The habitat type and the percentage of the hab- Profile Creek supports resident and fluvial bull itat covered by surface ice were recorded. Fish be- trout. Mature fluvial fish migrate from lower havior was recorded as resting (maintaining a posi- reaches of the South Fork Salmon and main Salmon tion on or above the substrate), hidden (beneath the rivers to spawn in Profile Creek. Other native fish substrate), or active (moving through the water col- include steelhead (Oncorhynchus mykiss), west umn). If I disturbed a fish, I did not search the area slope cutthroat trout (0. cZurki Zewisi), mountain where it reentered the substrate. Fish that darted whitefish (Prosopium williamsoni), dace (Rhinich- from cover before I was able to determine species, thys sp.), and sculpin (Cotfus sp.). focal point, or size were not included. I selected two study reaches (A, B) in the lower 5 I followed procedures outlined by Cunjak (1988) km of Profile Creek. Reach A was 88 m long with to describe microhabitat conditions. I labeled the an average width of 4.7 m; Reach B was 174 m long substrate under which the fish lay as the “home with a mean width of 5 m. I estimated habitat avail- stone” (Rimmer, Paim & Saunders 1984). Using a ability by classifying consecutive habitat types as wading rod and Marsh-McBirney current meter, I pool, riffle, run, or pocket-water (Bisson et al. measured water depth to the surface of the home 1982). Reach A contained 3 pools and 2 runs. Reach stone, focal point depth below the substrate surface, B contained 14 pools, 6 riffles, 2 runs, and 2 pocket- and focal point velocity. Focal point velocities were water habitats. Surface ice was present along stream measured after the home stone was removed. Water margins but did not span the channel; about 10% of temperatures were measured with a hand-held ther- the surface area of each habitat unit was ice cov- mometer at the focal point and 4 cm below the water ered. Anchor ice was uncommon and found prima- surface. I measured home stone dimensions, the dis- rily in riffles. tance from the home stone to the nearest shore, and the stream width perpendicular to the fish’s location. Material and methods I used a grid described by Overton, Chandler & Pisano (1994) to measure the percent of the sub- Before counting fish, I installed a continuously re- strate less than 2 mm at the fish’s location. cording thermograph to measure hourly water tem- I evaluated the association of each fish to cover peratures. Using a drysuit, mask, snorkel and re- above the substrate. Cover was defined as surface cording sleeve, I made day and night underwater turbulence, woody debris, boulders, riparian canopy surveys in each reach on 19-20 November 1991. within 1 m of the water surface, undercut banks, or Dives began at the downstream end of the lowest ice (Cunjak & Power 1986a). A fish was considered habitat in each reach. I proceeded slowly upstream under cover if one of these components was directly through each habitat searching for fish in the water overhead. column (Thurow 1994). I searched all areas of the I made night counts after day counts using the channel above the substrate, including beneath same procedures between 2200 and 0200 h. A hand- woody debris, undercut banks, or ice. Daytime sur- held halogen light was used to detect fish. When a 2 Winter bull trout fish was encountered, its species, size, and behavior No more than one fish was observed beneath a were recorded. At night, fish remained oriented to single home stone. When uncovered, bull trout typi- the substrate and I used the same technique to esti- cally remained motionless for 1 to 3 seconds, allow- mate fish size. I did not measure night microhabitat ing me to observe the focal point and estimate fish characteristics and did not attempt to locate fish hid- size.