North American Journal of Fisheries Management 25:1073±1081, 2005 q Copyright by the American Fisheries Society 2005 [Management Brief] DOI: 10.1577/M04-132.1 Extent of Anadromy in Bull Trout and Implications for Conservation of a Threatened Species SAMUEL J. BRENKMAN* AND STEPHEN C. CORBETT National Park Service, Olympic National Park, 600 East Park Avenue, Port Angeles, Washington 98362, USA Abstract.ÐThe use of radiotelemetry revealed that ulations that spend their entire life cycle in fresh- adult bull trout Salvelinus con¯uentus moved between water (Goetz 1989; Swanberg 1997). freshwater and the Paci®c Ocean and between water- One critical unanswered question is the extent sheds along coastal Washington. Forty-seven of 82 radio-tagged bull trout implanted in the Hoh River and to which bull trout move between freshwater and Kalaloch Creek basins were anadromous, and 23 ®sh marine environments. The extent of anadromy in were relocated in ®ve different estuaries up to 47 km bull trout remains uncertain primarily because ob- south of their tagging locations. Anadromous bull trout servations of native char have occurred in tidally typically entered the Hoh River from May to July, as- in¯uenced rivers in regions of overlap with the cended as far as 78 km upstream to spawning areas by closely related and morphologically similar Dolly September, and returned to the ocean to overwinter. This Varden, and the two species have not been distin- is the ®rst study to verify that anadromy is a primary life history form in coastal bull trout. The presence of guished during those observations (McPhail and anadromy and the irregular periods of freshwater, es- Baxter 1996). To our knowledge, the only pub- tuarine, and marine residence observed in this study are lished evidence that bull trout occur in marine wa- consistent with observations reported for other Salveli- ters is based on museum specimens that were col- nus species. These results also demonstrate that anad- lected in nearshore marine environments of the romous bull trout inhabit a diverse range of estuarine, Puget Sound, Washington, and the Fraser River, freshwater, and marine habitats. Their af®nity to estu- British Columbia (Cavendar 1978; Haas and aries and their observed temporal movements suggest that bull trout optimize winter refugia and forage op- McPhail 1991). portunities and are susceptible to incidental mortality in A prerequisite for successful conservation and gill-net and recreational ®sheries directed at other Paci®c recovery of bull trout populations is a more com- salmonids. An understanding of the extent of anadromy, plete understanding of life history or migratory sources of mortality, and diversity of habitats occupied forms and the extent of anadromy in bull trout. by bull trout is informative in the establishment of ap- The recognition of life history variability is im- propriate conservation and recovery strategies for this portant for maintaining biodiversity at the species species. level, and loss of such diversity may reduce the long-term ability of a species to adapt to changing The salmonid genus Salvelinus is known to ex- environments (Gresswell et al. 1994). It is known hibit diverse life histories that include anadromous that bull trout use an array of habitats in freshwater, and nonanadromous forms within a given species. but the degree to which they use marine habitats Arctic char Salvelinus alpinus, brook trout S. fon- is unknown. In this study, we used radiotelemetry tinalis, and Dolly Varden S. malma all exhibit com- to determine movements of adult bull trout im- plex migratory patterns and are known to be anad- planted with radio transmitters in the Hoh River romous (White 1941; Armstrong 1974; Nordeng basin and Kalaloch Creek on the Olympic Pen- 1983). The bull trout S. con¯uentus, a species of insula, Washington (Figure 1). The primary ob- char that has declined in distribution and abun- jectives of the research were to (1) determine dance throughout its range, is primarily con®ned whether and to what extent bull trout exhibit an- to interior drainages throughout northwestern adromy and (2) characterize spatial and temporal North America and is known to migrate exten- movements of bull trout in the Hoh River basin. sively in freshwaters. Recent studies have neces- We also discuss conservation implications related sarily focused on life histories of bull trout pop- to the observed patterns of movement and existing recreational and commercial ®sheries in coastal Washington. * Corresponding author: [email protected]. Study Area Received August 4, 2004; accepted November 24, 2004 The study area where bull trout were captured Published online July 20, 2005 and tagged included the Hoh River, South Fork 1073 1074 BRENKMAN AND CORBETT FIGURE 1.ÐMap of the study area depicting the tagging locations (solid circles) of 82 adult bull trout, the locations of ®ve ®xed telemetry stations (open circles), and the extent of aerial telemetry surveys in the Hoh River (rkm 0±85) and South Fork Hoh River (rkm 0±23) and along the Paci®c coast from Cape Flattery to Grays Harbor, Washington. Hoh River, and Kalaloch Creek. The Hoh River is sopium williamsoni, shorthead sculpin Cottus con- a glacial river that freely ¯ows in a westerly di- fusus, torrent sculpin C. rhotheus, reticulate scul- rection from its headwaters on the western slopes pin C. perplexus, coho salmon Oncorhynchus kis- of the Olympic Mountains to its con¯uence with utch, spring±summer and fall Chinook salmon O. the Paci®c Ocean (Figure 1). Sixty-®ve percent of tshawytscha, chum salmon O. keta, pink salmon the Hoh River watershed (a 45-river-kilometer O. gorbuscha, summer and winter steelhead O. my- [rkm] reach) occurs in Olympic National Park and kiss, sockeye salmon O. nerka, cutthroat trout O. is managed as a natural area. The lower portion of clarkii, longnose dace Rhinichthys cataractae, and the river ¯ows through state, tribal, and private Paci®c lamprey Lampetra tridentata. In 1994, the lands. The river drains 644 km2 and descends in National Park Service and the Washington De- elevation from 1,216 m in the headwaters to its partment of Fish and Wildlife implemented catch- con¯uence with the ocean. River discharge in the and-release regulations for bull trout in recrea- Hoh River basin is strongly in¯uenced by rainfall tional ®sheries. in autumn and winter and glacial runoff in spring and summer. Methods The South Fork Hoh River is the major tributary Radiotelemetry procedures.ÐThis study was de- to the Hoh River. The South Fork Hoh River drains signed to track movements of adult bull trout 130 km2 and descends in elevation from 1,475 to (.400 mm in total length) ®tted with individually 128 m. Kalaloch Creek is located 17 km imme- coded radio transmitters (Model MCFT-3A, Lotek diately south of the Hoh River, drains 45 km2, Wireless). Species identi®cation of all tagged ®sh and descends in elevation from 320 m to its con- was con®rmed by microsatellite analysis con- ¯uence with the ocean (Figure 1). Stream dis- ducted by the Wild Trout and Salmon Genetics charge is in¯uenced by rainfall during winter Laboratory at the University of Montana. Eighty- months. two bull trout were captured by angling and sur- In addition to bull trout, ®shes that inhabit the gically implanted with radio transmitters in the Hoh River basin include mountain white®sh Pro- Hoh River (n 5 58), South Fork Hoh River (n 5 MANAGEMENT BRIEF 1075 15), and lower Kalaloch Creek (n 5 9) (Table 1). and Kalaloch Creek were bull trout and not Dolly Fish were captured from July 2002 to June 2003 Varden (P. Spruell, University of Montana, un- at various locations throughout the study area, in- published data). Radio-tagged bull trout averaged cluding rkm 0±3 in the Hoh River estuary, rkm 507 mm in total length (range 5 404±695 mm), 27±65 in the Hoh River, rkm 0±11 in the South had a mean weight of 1.6 kg (range 5 0.8±3.6 kg), Fork Hoh River, and rkm 0±2 in Kalaloch Creek and were tracked for 10±20 months. (Figure 1). Captured bull trout were anesthetized Forty-seven of 82 (57%) bull trout moved from in a solution of tricaine methanesulfonate (80 mg freshwater to salt water and were considered to be per liter of water) for up to 7 min prior to deter- anadromous (Table 1). Of the remaining 35 ®sh mination of total length (mm) and weight (g). Each that were considered nonanadromous during their transmitter was tested for functionality before be- tracking histories, 21 moved exclusively in the ing inserted into the abdominal cavity in a manner Hoh River basin, three were considered dead and similar to that described by Ross and Kleiner the tags were later recovered, and 11 had tags that (1982). Radio transmitters were active for up to were assumed to have failed based on loss of con- 761 d, emitted a signal every 5 s, weighed 16 g tact. in air, and averaged less than 2% of ®sh weight. Upon completion of surgery, ®sh recovered in a Movements of Bull Trout Tagged in the Hoh cage in the river and were released within 20 min River Basin near their point of capture. All released ®sh ap- A total of 39 (of 73) bull trout that were tagged peared to exhibit normal behaviors upon release. in the Hoh River basin (including the Hoh River Radio signals did not transmit while ®sh occupied estuary and the South Fork Hoh River) moved salt water but resumed transmission when ®sh re- downstream and entered the Paci®c Ocean during entered an estuary and freshwater. various months. Some of these ®sh were relocated Tracking of ®sh.ÐThe tracking and relocation in coastal drainages located 5±47 km to the south of radio-tagged ®sh were accomplished by means of the Hoh River, including Cedar Creek (n 5 3 of ®ve ®xed stations distributed along riverbanks ®sh), Kalaloch Creek (n 5 1), the Queets River (n of the Hoh River basin (Figure 1).