Environ Biol Fish (2013) 96:691–700 DOI 10.1007/s10641-012-0059-5 Context-dependent diel behavior of upstream-migrating anadromous fishes Matthew L. Keefer & Christopher C. Caudill & Christopher A. Peery & Mary L. Moser Received: 31 October 2011 /Accepted: 18 June 2012 /Published online: 4 July 2012 # Springer Science+Business Media B.V. 2012 Abstract Variability is a hallmark of animal behavior during most of the diel cycle in low-gradient, and the degree of variability may fluctuate in response to less hydraulically complex reservoir and riverine environmental or biological gradients. For example, diel habitats. Movement shifted to predominantly diurnal activity patterns during reproductive migrations often (salmonids and American shad) or nocturnal (Pacific differ from those in non-breeding habitats, reflecting lamprey) at hydroelectric dam fishways where hy- trade-offs among efficient route selection, reproductive draulic complexity and predator density were high. phenology, and risk avoidance. In this study, we tested Results suggest that context-dependent behaviors are the hypothesis that diel movements of anadromous common during fish migrations, and that diel activ- fishes differ among freshwater migration habitats. We ity patterns vary with the degree of effort or preda- analyzed diel movement data from ~13 000 radio-, PIT-, tion risk required for movement. and acoustic-tagged adult fishes from five Columbia River species: Chinook salmon, Oncorhynchus tshawyt- Keywords Behavioral plasticity. Migration . scha; sockeye salmon, O. nerka; steelhead, O. mykiss; Orientation . Predation risk . Sensory ecology Pacific lamprey, Entosphenus tridentatus;andAmerican shad, Alosa sapidissima. All five species were active Introduction M. L. Keefer (*) : C. C. Caudill Department of Fish and Wildlife Sciences, Behavioral flexibility allows animals to rapidly and College of Natural Resources, University of Idaho, adaptively respond to changing conditions and to en- Moscow, ID 83844-1136, USA vironmental stimuli that signal resources or risk. Diel e-mail: [email protected] behavior, for example, varies widely along a continu- M. L. Moser um from almost exclusively nocturnal, to crepuscular, Northwest Fisheries Science Center, to strongly diurnal. Within species, diel rhythms can be National Marine Fisheries Service, quite plastic across life stages and at a variety of tempo- 2725 Montlake Boulevard East, ral scales in response to proximate and ultimate factors Seattle, WA 98112, USA (Helfman 1983; Boujard and Leatherland 1992; Reebs Present Address: 2002). Diel activity varies in response to environmental C. A. Peery conditions (e.g., photoperiod, water temperature), phys- Idaho Fishery Resource Office, iological status (e.g., starvation, metamorphosis), pred- U.S. Fish & Wildlife Service, 4147 Ahsahka Rd, ators, and life history requirements (e.g., migration, Ahsahka, ID 83520, USA reproduction). Ecological context additionally affects 692 Environ Biol Fish (2013) 96:691–700 daily rhythms, as individuals attempt to minimize 2007; Stansell et al. 2010). Our principal hypothesis was mortality risks while maximizing growth and fit- that fish movements would be temporally constrained in ness (Metcalfe et al. 1999; Railsback et al. 2005; high velocity and turbulent environments and areas of Aarestrup et al. 2009). increased predation risk. We further expected taxon- Diel behavior often significantly changes during specific responses that would correspond with differ- animal migrations, particularly for species that en- ences in sensory ecology among the study species. counter environmental and ecological conditions that diverge from those in rearing or other life history- specific habitats. Many neotropical birds, for instance, Material and methods are typically crepuscular or diurnal but undertake long-distance migrations solely at night (Gwinner Study species 1996). Similarly, diadromous fishes moving between freshwater and marine habitats often have very differ- The five study species exhibit a range of migration ent diel rhythms than those expressed in other life strategies and phenologies. The three salmonids are history stages (Smith and Smith 1997; Metcalfe et al. obligate migrants with high natal-site fidelity. Semelpar- 1998). Diel activity during fish migration is mediated ous Chinook salmon (spring-summer and fall-run) and by the sensory cues present in often dynamic environ- sockeye salmon follow relatively strict spawning migra- ments that can vary along several gradients that affect tion schedules. Summer steelhead can be either semel- the biophysical costs of movement and orientation or parous or iteroparous and the timing of their spawning perceptions of risk. These include hydraulic and visual migration is more variable. All three salmonids pass complexity (e.g., velocity, turbulence, turbidity), bio- through the Columbia River migration corridor from logical context (e.g., presence of predators, competi- spring through late fall and spawn in the year of river tors, or prey), and requirements associated with entry (Chinook and sockeye salmon) or the following orientation and navigation cues (e.g., visual environ- spring (steelhead) (Brannon et al. 2004). Our samples ment, olfactory plumes, temperature gradients, or so- included two Chinook salmon life history types: stream- cial interactions). These factors can elicit increasingly type salmon return as adults from March to July nocturnal or diurnal behavior depending on perceived (spring–summer Chinook) and ocean-type salmon re- risk-reward tradeoffs. Nocturnal migration may reduce turn from July to October (fall Chinook). American shad predation risk, for instance, but can exact fitness costs are mid-summer, iteroparous migrants whose natal site if migrants miss favorable migration windows or ar- fidelity is geographically coarse (Hendricks et al. 2002) rive late at spawning areas (Einum and Fleming 2000; compared to salmonids, whereas Pacific lamprey are Drent et al. 2003; Dickerson et al. 2005). semelparous, facultative migrants with no apparent natal Our primary objective was to assess diel migration homing (Goodman et al. 2008). Pacific lamprey migrate activity of five anadromous fish species moving up- in summer and fall, overwinter, and then spawn the stream through habitats differing in environmental com- following spring. The salmonids and lamprey are lith- plexity and passage difficulty in the Columbia River ophilic spawners that build and defend redds in main basin, USA. The study species were three native salmo- stem and tributary substrates, while shad are pelago- nids (Chinook salmon, Oncorhynchus tshawytscha; philic broadcast spawners in main stem reaches (Balon sockeye salmon, O. nerka; and steelhead, O. mykiss), 1975). a native lamprey (Pacific lamprey, Entosphenus tridentatus), and a non-native clupeid (American shad, Fish tagging and monitoring Alosa sapidissima). Diel activity was assessed at a suite of relatively low gradient riverine sites and at high Study fishes were primarily collected at Bonneville Dam gradient environments with evolutionarily-novel hy- (Fig. 1) at a trap facility adjacent to the north-shore draulic and physical elements created by large hydro- fishway (salmon, steelhead, shad) or using passive box electric dams. The dams and recent aggregations of traps affixed to fishway walls (Pacific lamprey). Less large predators in some dam tailraces and fishways have than 5 % of the tagged shad were captured by angling been associated with upstream migration delay and ele- or electro-fishing in the Bonneville Dam tailrace. Details vated mortality risks (Moser et al. 2002; Caudill et al. of salmonid and Pacific lamprey capture, anesthetization, Environ Biol Fish (2013) 96:691–700 693 Dallas, Texas) and American shad in 2005–2007 (full- duplex tags; ~2-mm×12 mm; Texas Instruments). All of these fishes were released at sites downstream or up- stream from Bonneville Dam and monitoring occurred after recovery and natural resumption of upstream movements. Two groups of fish were collected and radio-tagged at lower Snake River dams: transmitters (8.3-mm×18.3-mm, 2.1 g in water; Lotek Wireless Inc.) were surgically implanted in Pacific lamprey in 2006– 2008 (McIlraith 2011) and were intragastrically inserted (1.1-cm×4.3-cm, 7.7 g in air; Lotek Wireless Inc.) into sockeye salmon in 2000 (Keefer et al. 2008). These fishes were released upstream from Lower Granite Dam (Fig. 1). Fig. 1 Map of the Columbia River, U.S.A., basin showing sites where diel activity of tagged adult Chinook salmon, steelhead, Fifteen monitoring sites were used in the diel evalua- sockeye salmon, Pacific lamprey, and American shad was mon- tion (Fig. 1). These were nested within five habitat types itored. Upstream migrants were collected at Bonneville Dam that varied qualitatively along environmental and ecolog- and at lower Snake River dams. Activity was monitored at nine ical gradients: (1) unimpounded tributaries (Klickitat, dam sites, which included the tailraces, fishways entrances, and fish ladder exits at Bonneville, The Dalles, and John Day dams, Deschutes, and Clearwater rivers, Fig. 1); (2) large main and at six riverine and reservoir habitats shown on the map as: stem reaches (Bonneville reservoir, Snake River above (1) Bonneville reservoir; (2) Klickitat River; (3) Deschutes Lower Granite reservoir, and Hanford Reach of the Co- River; (4) Columbia River Hanford Reach; (5) Clearwater Riv- lumbia River); (3) dam tailraces; (4) fishway entrances; er; and (6) Snake