1.1. Anadromous salmonid biology

Although quite variabile in their life histories (within and among species), anadromous salmonids share the following traits. Adult fish spawn in freshwater streams or lakes, usually in late summer or fall (Groot and Margolis, 1991). Their large yolky eggs are buried in the substrate, and embryonic development occurs here (Thorpe, 1984). The juveniles emerge from the substrate the following spring as "fry" and are dependent on external food sources upon emerging (Thorpe, 1984). The life histories of the various species diverge at this point, with some species migrating to the estuary at this stage and other species delaying their migration for months or years (Northcote, 1984). After passing through the estuary, the fish carry out most of the growth in the ocean. Depending on the species and stock, the fish spend between one and seven years in the ocean (Groot and Margolis, 1991). Adults then return to their natal streams or lakes (although some straying is common (Quinn, 1984)) and die shortly after spawning.

chinook salmon

Chinook salmon are divided into two "races" (or subspecies, depending on nomenclature), both of which inhabit the Columbia River system. "Ocean-type" chinook return as adults in the late summer or fall and spawn almost immediately after reaching the natal stream (Healy, 1991). The juveniles migrate as subyearlings, usually several months after emerging as fry, although timing of emigration is quite variable (Reimers and Loeffel, 1967). This group is also referred to as "chinook 0's" or as fall chinook. Ocean-type chinook are generally found in the southern part of the species' range. "Stream-type" chinook return as adults in the spring and delay spawning for several months. The juveniles migrate as yearlings after overwintering in the river environment. These fish, also referred to as "chinook 1's" or as spring chinook, are generally found in the northern part of the species' range. Although the two types of chinook may occupy the same streams, they appear to be genetically distinct (Carl and Healy, 1984) and show heritable behavioral differences (Taylor and Larkin, 1986; Taylor, 1988). Stream-type juveniles display higher levels of antagonistic behavior and stronger positive current response, consistent with defending territory and extended residence in streams.

sockeye

The life history of sockeye salmon is the most variable of all the Pacific salmon, with a wide variety of adaptations for specialized conditions (Burgner, 1991). In addition to the anadromous form, there is a landlocked form commonly referred to as kokanee. Anadromous sockeye usually spawn in the tributaries of lakes (Groot, 1982). Upon emergence, the fry migrate to a nursery lake where they may spend 1 to 3 years. The sockeye smolts then migrate downstream to the ocean. Ocean residence for sockeye is variable, ranging from 1 to 4 years (Burgner, 1991).

steelhead trout

Steelhead trout (Oncorhynchus mykiss) is the same species as rainbow trout, with steelhead a migratory form and rainbows a landlocked form. Steelhead, until recently, were classified as Salmo gairdneri, partially reflecting their morphological and behavioral similarities to Atlantic salmon (Salmo salar) (Netboy, 1980). The change of nomenclature is based on the Pacific coast origin of the species and an alignment with Pacific salmon (Light, et al., 1989). The Columbia River Basin is the world's largest producer of steelhead (Netboy, 1980; Light, 1987). Steelhead are generally split into two races: "winter" steelhead return as adults between November and April; and "summer" steelhead return as adults from May to October (Withler, 1966). In the Columbia Basin, winter-run steelhead are found exclusively west of the Cascades, while summer-run steelhead are found in some western tributaries and are the only steelhead found in the Snake and upper Columbia Rivers and their tributaries (Pevin, 1990). Smolts usually migrate in the spring of their second year, but there is variability in the duration of freshwater residence (Withler, 1966). The majority of steelheads spend 2 years in the ocean before returning as adults (Pevin, 1990). Unlike Pacific salmon, steelhead don't always die after spawning (Childerhouse and Trim, 1979). A small percentage return to the ocean after spawning and then return back to freshwater the following year to spawn again.

coho

In Washington and Oregon, coho are found primarily in coastal streams and tributaries of the Lower Columbia (Sandercock, 1991). The freshwater residence of coho is quite variable, and they have the most extended stream residence of Pacific salmon (Taylor and Larkin, 1986). Because few wild populations of coho undergo extensive migrations in the Columbia River or its tributaries, I do not analyze any coho data in this thesis.

smoltification

The initiation of migration is preceded by the parr-smolt transformation (smoltification) (Folmar and Dickhoff 1980), in which the juveniles transform from a stage in their life history adapted for stream inhabitation to a stage adapted for downstream migration and eventually saltwater inhabitation. Smoltification is a series of morphological, physiological, and behavioral changes. A discussion of smoltification is important for two reasons. First, the morphological, physiological and behavioral changes are all related, and thus understanding how each operates can help elucidate the behavioral changes important to modeling. Also, it is clear that the timing associated with smoltification is critical, and this lends importance to the travel time studies.

Behaviorally, the fish undergo several changes. Prior to smoltification, the fish exhibit positive rheotaxis (Thorpe and Morgan, 1978), and maintain their position in the river or lake. They are also territorial bottom dwellers. Upon smoltification, fish are less prone to hold position against the current, and thus downstream movement becomes initiated. In addition, they become less territorial and more surface oriented.

Morphological changes that occur during smoltification are a silvering in body color and a decrease in weight per unit length (commonly referred to as condition factor) (Wedemeyer, et al. 1980), resulting in a more slender and streamlined fish. Some evidence exists for a threshold size that may be important in the timing of seaward migration (Folmar and Dickhoff, 1980).

Physiologically, several changes occur during smoltification. First, there is heightened hypoosmotic regulatory capability that increases salinity tolerance and preference. Endocrine activity increases, notably in greater levels of thyroxine, and according to Hoar (1965), the endocrine system forms a chemical link between the organism and the environment. The higher hormonal levels may also induce a behavioral response; Godin et al. (1974) demonstrated that artificially increasing thyroxine levels in Atlantic salmon smolts leads to increased migratory behavior. Also, an increase in gill Na⁺-K⁺ ATPase activity is typical of fishes existing in saltwater environments. In fact gill Na⁺-K⁺ ATPase is often sampled to assess the level of smoltification in juveniles (Zaugg, 1982).

Clearly, smoltification is a complex process, and events are coordinated such that fish are ready to enter saltwater at the appropriate time. Flagg and Smith (1982) determined that juvenile coho with visuals signs of smoltification suffered no loss of swimming stamina when transferred from freshwater to a seawater, while juveniles without these signs did suffer a loss in swimming stamina. Fish that weren't transferred from fresh water to sea water at the proper time appeared sluggish, potentially increasing their susceptibility to predation. Flagg and Smith (1982) also determined that mortality associated with salt water stress was is inversely related to levels of thyroxine and Na⁺, K⁺ ATPase, which are indicators of degree of smoltification. Observations also show that some species of salmonids revert back to a freshwater adapted state if they don't reach saltwater within a certain time frame (Hoar, 1976). It appears that a species and stock specific optimal period for reaching saltwater exists that maximizes survival of the fish. Thus, modeling the temporal aspects of migratory behavior can be beneficial in coordinating migrations of hatchery stocks and in determining deleterious effects of delaying the migration of wild stocks.

juvenile salmon migratory behavior

Clearly, many facets of juvenile salmon migratory behavior are not well understood. Behavior patterns are quite variable among species, and in some cases, among stocks. It possible to generalize some types of behavior across species, but with other types of behavior it is important to note differences. In many cases where a group of workers establishes a behavior pattern for a particular species, another group offers a counter example

In this section, I present some questions pertaining to salmon migratory behavior and results of studies examining these issues. While my focus is on the behavior of steelhead, chinook, sockeye, and coho, I will also present results based on other species of anadromous salmonids, including Atlantic salmon (Salmo salar) and sea trout (Salmo trutta).

• What cues the initiation of migration?

• Is migration active or passive?

Also consistent with passive migration in coho and Atlantic salmon are studies determining that some smolts exhibit a loss of swimming proficiency as compared to fish in the parr stage. Smith (1982), Flagg and Smith (1982) and Glova and McInerney (1977) observed this with coho, and Thorpe and Morgan (1978) determined sustained swimming velocities of Atlantic salmon juveniles decreased substantially during the period of peak downstream migration. It is not clear, though, whether this loss of swimming "proficiency" is due to a physical change or, as Thorpe and Morgan (1978) speculate, "a behavioral refusal to undergo sustained swimming."

• Do migrating juvenile salmonids have distinct diel patterns?

• Do fish school as they migrate?

• What factors influence migration rate?

• What is the spatial distribution of fish in the river?