Effects on Predation and Prey-Refuge Habitat

Shorezone structures are expected to affect predation on ESA-listed salmonids by simplifying the shoreline (bulkheads eliminate shallow water, complex woody debris, overhanging vegetation, and complex substrate, and create energetically unfavorable shoreline conditions); providing foraging habitat (shade and overhead cover) for structurally-oriented ambush predators, specifically largemouth and smallmouth bass (piers); and by extending the duration of predation by allowing visual predators to forage at night (piers with artificial lighting). Due to their life history strategies, largemouth and smallmouth bass are the predators most likely to benefit from shorezone structures; additional discussion of the ecology of largemouth and smallmouth bass is warranted (see "Bass" section above), as well as a brief discussion of general predator-prey interactions. However, cutthroat trout are the primary predator of salmonids in the nearshore from February through June, and in the offshore for the rest of the year (Warner, pers. comm., 7 July 2000). The effects of shorezone structures on the efficiency of cutthroat predation on salmonids are unknown. It is likely that the loss of complex refuge habitat resulting from shorezone development would represent a disadvantage to juvenile salmonids in the presence of mobile predators such as cutthroat trout.

Predator-prey Interactions

For juvenile salmonids, the net loss in complex cover resulting from the replacement of natural shorelines with docks and bulkheads may be critical. Historically, the littoral zone of Lake Washington contained abundant aquatic and shoreline vegetation, and woody debris (Evermann and Meek 1897; Stein 1970). The lowering of the water level and substantial shoreline development have eliminated much of the shallow-water habitat available to juvenile salmon. Docks and piles may provide shallow-water cover for juvenile salmon, but they also provide cover for bass. Cooper and Crowder (1979) stated that "reducing structural complexity may remove prey refuges and subject the remaining prey to high risk until they are decimated." Docks, piles, and bulkheads are relatively simple structural elements compared with rootwads and trees with branches, and other forms of natural cover found along undisturbed shorelines.

Sustainable predator-prey interactions in general require the existence of prey refuge to prevent the extermination of the prey organism. Numerous studies have reported increased use of complex cover (e.g., aquatic vegetation, woody debris, substrate interstices, and undercut banks) by prey fishes in the presence of predators, and reduced foraging efficiency of predators due to habitat complexity (e.g., Bugert and Bjornn 1991; Persson and Eklov 1995; Werner and Hall 1988; Tabor and Wurtsbaugh 1991; Wood and Hand 1985). Savino and Stein (1989) demonstrated that refuge is critical for prey fish survival; their study found that largemouth bass captured all prey fish that strayed from areas with aquatic vegetation into open water. Bass also eliminated all prey fish from pools that provided no refuge in a study by Schlosser (1987), while predator and prey were able to coexist in pools with complex cover. Hixon and Beets (1993) provided evidence of the value of complexity in a study of marine reef fish; prey fish were most abundant on reefs where refuge size closely matched the body size of the prey species, and where the number of refuge holes was not limiting. Lynch and Johnson (1989) showed similar results for juvenile bluegill (Lepomis macrochirus) in fresh water. Gotceitas and Colgan (1989) found that prey fish in fresh water preferentially selected refuge habitat with greater complexity than was necessary to significantly reduce foraging success of predators. Helfman (1979) suggested that the utilization of small floating objects on bright days by prey fish was related to the visual advantage the prey fish gained by being shaded over a predator approaching from the brightly lit surrounding area.

Shallow water functions as a refuge from predation for small fish, especially in the absence of complex habitat features such as woody debris or submerged vegetation. In Schlosser's study (1987), bass eliminated prey fish from structurally simple pools either by direct consumption, or by forcing the prey fish into shallow-water habitats, thus subjecting prey fish to potentially decreased feeding opportunities. Bass predation also excluded grazing minnows from all but the shallow sections of pools in Oklahoma streams studied by Power et al. (1985). Ruiz et al. (1993) reported that mummichogs (Fundulus heteroclitus)(<51 mm) in a subestuary of Chesapeake Bay preferentially occupied shallow water (<35 cm) in the absence of submerged aquatic vegetation. Collins et al. (1995a) found that feeding rates by small fish (<100 mm) in two Ontario shield lakes were 10 times higher in shallow water (<20 cm) than in the rest of the littoral zone. Littoral slope has been negatively correlated with fish numbers and positively correlated with fish size (Randall et al. 1996). Brown (1998) observed no piscivores in "littoral fringe" (within 2.5 m of shore) transects in Lake Joseph, Ontario.While most of the above studies on predator-prey interactions were from warmwater systems, studies of juvenile salmonid response to predators are analogous. Juvenile salmonids modify their behavior in the presence of predators by seeking or orienting to complex refuge (Gregory and Levings 1996; Reinhardt and Healey 1997), emigrating from areas with predators (Bugert and Bjornn 1991), aggregating (Tabor and Wurtsbaugh 1991), and adopting diel vertical migrations (Eggers 1978). The response of juvenile salmonids to predators increases with experience (Healey and Reinhardt 1995) and body size (Reinhardt and Healey 1997). Behavioral responses can be influenced by environmental factors such as visibility. Turbidity reduces predator-avoidance behavior in salmonids (Gregory and Levings 1996; Gregory 1993), and reduces prey mortality rates by reducing the prey-encounter rates of predators (Ginetz and Larkin 1976; Gregory and Levings 1998; Beauchamp et al. 1999). Salmonid predators also modify their behaviors in response to habitat complexity. Piscivorous brook trout in Quebec lakes switched foraging tactics from active cruising to ambushing when prey refuges were present (East and Magnan 1991).

Simplification of shoreline habitat, reducing the availability of prey refuge-habitat, should be avoided. Predator-prey interactions modify the behavior of both predator and prey species. Prey refuges are essential for the continued existence of vulnerable prey species. Complex habitat features that exclude predators, physically or through risk-aversion, can function as prey refuge. Examples of effective prey refuge may include shallow water, complex substrate, aquatic and emergent vegetation, overhanging terrestrial vegetation, undercut banks, and woody debris. Efforts to restore habitat function along lakeshores should be encouraged.

Shorezone Structures

Separating the effects of shorezone structures on juvenile salmon into discussions of the effects of individual structures in isolation may not yield the most appropriate conclusions since development seldom occurs as an isolated structure. The effects of shoreline development in its entirety should also be included in the discussion. Jennings et al. (1999) stated that "fish do not respond to shoreline structures: rather, they respond to a suite of habitat characteristics that are the result of the structure, changes to the riparian zone associated with its placement (vegetation and woody structure removal), and often, intensive riparian zone management that occurs on developed properties." Brazner (1997) found that sites adjacent to human development in Green Bay, Lake Michigan had fewer fish and species, and had more disturbance-tolerant fish assemblages. Fish species richness and abundance were highest in undeveloped wetland habitats (Brazner 1997). Species richness and total fish abundance were less at developed sites than at undeveloped sites in the littoral zone of Spirit Lake, Iowa (Bryan and Scarnecchia 1992). Poe et al. (1986) found that an undeveloped bay was characterized by a percid-cyprinid-cyprinodontid assemblage, while a developed bay (bulkheaded shoreline, frequent dredging, low macrophyte species richness, reduced water quality) was dominated by a centrarchid (bass, sunfish) assemblage. Both Poe et al. (1999) and (Bryan and Scarnecchia 1992) found that fish species richness was positively correlated with macrophyte species richness. Lange (1999) provided evidence that residential shoreline development is "a likely agent in causing system-wide disruption to fishŠ." Sites with combinations of development structures (i.e., dock and bank stabilization) had low fish abundance and richness (Lange 1999). Lange (1999) generally concluded that the results of cluster analysis indicated that "sites associated with high occurrence of all forms of development and low occurrence of vegetation, tended to have the lowest total abundance and species richness, regardless of observational scale." Both Jennings et al. (1999) and Lange (1999) found that the scale of one's observations affects conclusions, and the cumulative impacts of multiple development features may be substantial. With at least 81 percent of Lake Washington shoreline bulkheaded and at least 2.5 percent and 4 percent of the shallow-water habitat covered with residential piers in Lakes Sammamish and Washington, respectively, the potential for cumulative adverse impacts is significant.It is within this context - that shoreline development in general degrades aquatic communities - that we examine the effects of individual structure types on those communities. Individual structure types often occur together, confounding inference about their respective impacts. Additionally, the extrapolation of results among systems can be uncertain due to the physical and biological differences between systems. However, a negative response to human disturbance and habitat alteration is consistent among diverse aquatic/marine communities.

Piers

As discussed above, bass utilize structural features (natural or artificial) for both foraging and spawning. Christensen et al. (1996) found a significantly negative correlation of lakeshore development with CWD in 16 lakes in northern Wisconsin and upper Michigan. Qualitative observations in Lake Washington indicate that little woody debris can be found along developed sections of the shoreline. Piers provide alternative sources of shade, overhead cover, and in-water structure (piles and boatlifts) that bass could utilize for foraging and spawning, in the absence of natural features. Observations by Stein (1970) and Pflug (1981) in Lakes Washington and Sammamish respectively, indicate that bass do occasionally occupy piers.

Freshwater: As expected, the literature review did not produce any studies of the relationship between piers and bass predation on juvenile salmonids. Studies from freshwater systems also lacked evidence for predator aggregations associated with piers. A study in Lake Washington found no significant differences in catch-per-unit-effort of any fish species between under-pier and control sites, but few bass were captured (White 1975). However, evidence for the use of piers by bass in Lakes Washington and Sammamish was located. Stein (1970) reported that largemouth bass were commonly found under piers in Lake Washington during the spring, but considered natural cover to be their preferred habitat. Unpublished results of a study by the Muckleshoot Indian Tribe in Lake Sammamish indicated that smallmouth bass were preferentially locating nests proximate to residential piers (Malcom, pers. comm., 13 April 2000). Although residential piers only covered approximately 13 percent of the nearshore zone (0-20 m from shore), 32 percent of the smallmouth bass nests were within 2 meters of piers, and 54 percent were within 2 meters of a pier or other artificial structure (i.e., isolated piles, water pipes, boat launch rails, tires, rebar) (Malcom, pers. comm., 13 April 2000). Shade was apparently not a critical attraction feature of piers for spawning smallmouth bass; instead, the attraction was to physical structure provided by piers, further evidenced by the location of nests adjacent to non-shading structures such as isolated piles (Malcom, pers. comm., 13 April 2000). This finding does not indicate that shade was unimportant to foraging smallmouth bass, only that bass were not preferentially locating nest sites in shady locations. The findings of Malcom (pers. comm., 13 April 2000) corroborate the findings of Vogele and Rainwater (1975), who also found that smallmouth bass nests were not closely associated with sheltered habitat in Bull Shoals Reservoir. The majority of smallmouth bass nests were beside submerged stumps in gravel and rubble substrates, while largemouth bass nests were either under artificial brush shelters or adjacent to a submerged log, rock, or tree base (Vogele and Rainwater 1975).

Additional evidence for a connection between bass and piers comes from unpublished data. WDFW personnel electrofishing for bass in 50 to 70 local (western Washington) lakes observed that bass were more often associated with natural structures such as brush piles, beaver lodges, and overhanging willows and, to a lesser degree, were found under docks or adjacent to piles, but empirical evidence to support these observations was not collected (Bonar, pers. comm., 13 June 2000). Qualitative observations by Bonar (pers. comm., 13 June 2000) suggest that structures concentrate bass in lakes where structure is limiting. One-third of the largemouth bass in Lake Baldwin, Florida showed a significant preference for piers in the absence of aquatic vegetation (Colle et al. 1989).

Two studies (that did not include bass) of freshwater fish use of piers did not find evidence of predator aggregation. Ward et al. (1994) did not find a relationship between shoreline development (including piers) and northern pikeminnow predation on outmigrating chinook and steelhead in the lower Willamette River, Oregon. Northern pikeminnow were more abundant along undeveloped than developed reaches of the lower Willamette River (Ward et al. 1994). In a study in Lake Tahoe by Beauchamp et al. (1994), day and night patterns in fish density and species composition were similar between docks and open shoreline. Neither of these studies specifically investigated the relationship between piers and bass or cutthroat predation on salmonids.

Marine or Estuarine: Several studies from East Coast estuarine systems provided contradictory results that may reflect differences in systems and study designs. Low fish abundance and species richness under piers compared with pile field and open-water sites (Able et al. 1998), and low fish growth rates under piers compared with pier edges and open-water (Duffy-Anderson and Able 1999) have been reported in the Hudson River estuary. Conversely, in the Rhode River estuary, Toft et al. (1995) reported significantly greater abundance of several fish species under piers than 10 meters away. Local studies of estuarine systems that included salmonids were more consistent, indicating that juvenile salmonids forage under piers, and that predator aggregations were not observed. Ratté (1985) reported that juvenile chinook and coho salmon foraged under a large commercial pier in the Commencement Bay estuary, Washington, and that no aggregations of predators or selective predation on salmonids was observed. Juvenile chum and pink salmon were attracted to a large pier complex (submarine berth) during daylight at the U.S. Navy Bangor submarine base on Hood Canal, but aggregations of predators were not observed (Prinslow et al. 1979 and 1980). Findings from marine and estuarine systems should be considered to be the least applicable to the Lake Washington system.

Rock Crib Structures: Several studies have examined the effects of shoreline development in general on various indices of fish community structure (e.g., Poe et al. 1986; Brown 1998; Lange 1999; Jennings et al. 1999). Of these studies, Brown (1998) and Lange (1999), while examining the effects of shoreline development in general, included analysis of fish response to moorage structures in Lake Joseph and Lake Simcoe, Ontario, respectively. The moorage structures in these Ontario lakes differ somewhat from the typical structures found in local lakes. The majority of the moorage structures on Lake Washington are piers supported by piles (typically wood piles, 20-30 cm in diameter). The majority (> 85%) of the structures in Brown's (1998) study on Lake Joseph were crib structures, that is docks or boathouses supported by log cribs filled with boulders; only 8 percent were piers supported by piles. While not a typical design for residential piers in Lake Washington, these types of structures are common in the Ship Canal. Lange (1999) examined fish response to a variety of structure types at three different scales in Lake Simcoe, differentiating between structures supported by piles ("temporary docks") or cribs ("permanent docks"). Significant negative effects of temporary docks (most similar to our local piers) on fish richness or abundance in Lake Simcoe were not observed at any scale (Lange 1999). Crib-supported docks did have a significant positive effect on fish abundance at the two largest scales (244 m, 488 m), but not at the smallest scale (122 m) (Lange 1999). Brown (1998) also found that crib structures increased densities of forage fish (<100 mm) in the littoral fringe on exposed shorelines or in areas where CWD had been removed. Brown (1998) speculated that interstitial spaces within crib structures provided refuge from waves and predation for small fish along exposed shorelines. In Lake Tahoe, up to ten-fold higher densities and a greater diversity of small fishes were associated with rock-crib structures, whereas fish assemblages around pile piers did not differ from paired adjacent areas without shorezone structures (Beauchamp et al. 1994); however, bass and other centrarchids were absent from the main basin of Lake Tahoe at the time of this study. Brown (1998) observed (qualitatively) large numbers of piscivores beyond the littoral fringe around crib structures in Lake Joseph, and suggested that their presence was a response to the abundance of forage fish.

Conclusions: These findings, when considered with existing knowledge of bass ecology, suggest that bass prefer natural cover for foraging, and preferentially site nests adjacent to structures, but bass utilize piers, piles, and other artificial structures for foraging and nesting in lieu of natural cover or structure. Piers and piles differ from natural cover/structure elements such as brush piles, primarily in their lack of structural complexity. This difference is critical for prey fish, which rely on structural complexity for survival in the presence of predators, particularly mobile predators such as cutthroat trout. In developed lakes, piers become the dominant structural features at the expense of natural complex structures such as woody debris and emergent vegetation. That bass and other predators gain an advantage over prey fish in structurally simple environments is substantiated by findings that bass (especially smallmouth bass) persist or thrive along developed shorelines, while other species decline (Brown 1998; Bryan and Scarnecchia 1992; Poe et al. 1986; Lange 1999). Recognition of this advantage to bass and other predators necessitates a cautious scrutiny of proposed new and modified piers while awaiting results from the direct studies on the relationship between piers and other shorezone structures, and bass predation on salmonids that are currently underway (i.e., the studies of Roger Tabor, USFWS; Kurt Fresh, WDFW; Rod Malcom and Eric Warner, Muckleshoot Tribal Fisheries). Regardless of the development proposal, any project that would potentially reduce the structural complexity of the shorezone should be considered likely to adversely affect ESA-listed salmonids. New piers should also be considered as new, structurally simple habitat elements that provide cover and structure to spawning and foraging bass, and perhaps other predators. Replacement piers and pier modifications should be viewed by regulatory agencies as opportunities to regain some habitat function and minimize overwater coverage (see "Productivity" section below).

It is useful to ask what features of piers make them attractive to bass in the lacustrine environment. Male bass preferentially locate nests adjacent to structural features such as rocks or logs, apparently to reduce the perimeter that must be guarded or to provide visual isolation from nearby conspecifics (Heidinger 1975). Thus, for spawning bass, pier elements that protrude from the substrate (i.e., piles, boatlifts, etc.) may be attractive. The initial data suggests that this is the case for smallmouth bass (Malcom, pers. comm., 13 April 2000). The structure provided by piers and boatlifts may potentially increase spawning habitat and/or reproductive success of bass.

It is less clear what pier features primarily attract foraging bass: shade, overhead structure independent of shade, vertical structural elements, or a combination of features. The finding that largemouth bass were more likely to hover under large swimming floats than small study floats, and that fish were generally not observed under "sham" floats consisting of wood frames only, suggests that shading may be key, and that the dimensions of the area shaded may also be important (Helfman 1979). Anecdotal evidence from anglers supports the hypothesis of the importance of the dimensions of the overwater area, as fishing efforts are directed at the portions of piers with the most surface area (broad ells). Despite circumstantial evidence and an intuitive connection, direct evidence for a correlation between pier shade-production and bass occupation was not located in the course of this review. The circumstantial evidence does indicate a need for both further study and critical appraisal of pier design in the interim. Studies investigating the effects of light-transmitting devices (prisms, grating) on bass use of piers would also be useful. Prisms are currently being mandated as mitigation for overwater coverage by several municipalities on Lake Washington and Lake Sammamish. However, their ability to reduce bass attraction to piers has not been proven, despite their ability to transmit ambient light.

One additional note on piers in Lake Washington is the prevalence of illegal lake-water withdrawals. Many waterfront property owners have illegal pump systems for withdrawing water from Lake Washington. This may be a critical source of water loss during the dry season, and could ultimately affect fish passage at the Ballard locks.

Bulkheads

Studies of the relationship between shoreline armoring and predation on juvenile chinook or coho salmon in Lake Washington and Lake Sammamish were not found. Cautious conclusions about the effects of shoreline armoring on predation can be drawn from studies of predator-prey interactions and the habitat use by small non-salmonids in other north-temperate lakes, and studies of salmonid habitat use in large rivers and reservoirs. While no direct links were identified between predation and bulkheads, an intuitive connection exists between the loss of complex, shallow-water foraging habitat for juvenile salmonids and an increased exposure to potential predation. Bulkheads could directly affect predation on juvenile salmonids by eliminating shallow-water refuge habitat or, indirectly, by the elimination of shoreline vegetation and in-water woody debris that generally accompanies bulkhead construction. The importance of shallow-water refuge habitat and complex habitat features to small fish has been discussed above. Juvenile fall chinook salmon in the Columbia and Snake Rivers and (preliminary results suggest) Lake Washington have demonstrated a preference for shallow, low-angle shorelines, although the motivation for this observed preference has not been fully investigated (Key et al. 1994a and 1994b; Garland and Tiffan 1999; Tabor, pers. comm., 9 June 2000). Placing bulkheads waterward of ordinary high water (OHW) eliminates the shallow water identified by Collins et al. (1995b) as critical for foraging, refuge, and migration of small fish (> 100 mm). The simplification of the shoreline (i.e., removal of CWD and shoreline vegetation) that typically accompanies bulkhead construction (Christensen et al. 1996) further reduces refuge habitat. Lange (1999) found that bank stabilization (i.e., various forms of erosion control structures referred to as "bulkheads") was negatively correlated to fish abundance and species richness at all spatial scales investigated in Lake Simcoe, Ontario.

The finding that both fish species richness and abundance were negatively correlated with bulkheads at every scale (Lange 1999) indicates that fish in Lake Simcoe generally avoid bulkheads. Juvenile fall chinook in the Columbia and Snake Rivers were found to avoid riprap shorelines (Key et al. 1994a and 1994b; Garland and Tiffan 1999). Young-of-year bass in Lake Joseph, Ontario, did not exhibit a preference for spatially complex habitat in Brown's (1998) study, and may represent an exception to the avoidance theory. Jennings et al. (1999) found that species richness was greater along riprap bulkheads than smooth vertical bulkheads in 17 Wisconsin lakes. Riprapped shorelines in the study by Jennings et al. (1999) also had greater species richness than unarmored shorelines (does not imply natural, only the lack of armoring structure), but they cautioned that the findings were an artifact of the scale of the investigation, the heterogeneity of the unarmored sites, and the increased effort required to obtain estimates of species richness at unarmored sites. Converting lakes entirely to riprapped shoreline would ultimately reduce species richness at the lake scale, but in situations where hard shoreline armoring is necessary, riprap would be preferred over vertical walls for fish habitat (Jennings et al. 1999) (the specifications of the various shorelines compared by Jennings et al. 1999 are available in a report that was not obtained for this review - Jennings et al. 1996). However, Jennings et al. (1999) were not considering a situation where an endangered species could be potentially jeopardized by the shoreline protection method. Riprap may provide greater habitat heterogeneity and ultimately greater species richness than smooth vertical bulkheads, but the effects of habitat heterogeneity on predation were not investigated.The use of riprap shoreline protection in Lake Washington could provide concealment habitat to the most abundant native piscivores - cottids Tabor et al. (1998) reported predation on salmonid juveniles by sculpins greater than or equal to 50 mm in length. Few such sculpin were found over sand/mud substrates relative to gravel/cobble substrates in Lake Washington, due to the lack of refuge habitat in sand/mud substrate (Tabor et al. 1998). In Lake Washington and the Cedar River, cottid size was generally positively correlated with substrate size, and riprap shorelines had large cottids relative to sites with smaller substrate particles (Tabor et al. 1998). Bulkheads in Lakes Washington and Sammamish are typically nearly vertical, and constructed of large boulders with large interstitial spaces. The large interstitial spaces found within riprap shorelines provide concealment to abundant, large native cottids.

In summary, bulkheads eliminate shallow-water habitat and complex habitat features that may function as critical prey-refuge for juvenile chinook and coho salmon. Bulkheads have been shown to reduce the diversity and abundance of all fish species except smallmouth bass in other north-temperate lakes. Riprap bulkheads, which provide interstitial spaces that can be utilized by a variety of invertebrate and fish species, may provide refuge habitat for piscivorous sculpin, while also eliminating the shallow water refuge for juvenile salmonids. No evidence was found for positive effects of shoreline armoring on aquatic species.

Lighting

Studies of the effect of pier lighting, on predation of juvenile salmonids in lakes were not found in the course of this review. Western grebes have been observed foraging at night around artificial lights in Lake Washington (Tabor, pers. comm., 9 June 2000). Grebes, blue herons, and other birds have been observed feeding at night on the Cedar delta in the portion that is lit up by The Boeing Company lights (Warner, pers. comm., 7 July 2000). Prey behavior can influence light-mediated predation rates by both increasing exposure to predators by slowing migration rates through rivers, and reducing capture efficiency by increasing avoidance behavior. In freshwater laboratory experiments, Tabor et al. (1998) found that prickly and torrent sculpin were capable of preying on sockeye fry in complete darkness, but predation rate declined with increasing light intensity. An increase in predator avoidance ability by sockeye fry with increasing light intensity may explain this inverse relationship (Tabor et al. 1998). Petersen and Gadomski (1994) observed a similar relationship (decreasing predation rate with increasing light intensity) between northern squawfish (northern pikeminnow) and juvenile chinook salmon, and offered the same explanation, as did Howick and O'Brien (1983) for bass-bluegill interactions and Mazur (Univ. of Washington, unpubl. data) for juvenile trout responding to lake trout. Alteration (slowing) of migratory behavior and subsequent increased sculpin predation rates on sockeye fry with increasing light intensity were observed in simulated stream experiments (Tabor et al. 1998).The nocturnal behavior of juvenile chinook and coho and their predators in Lake Washington and Lake Sammamish, and their response to lighting is poorly understood. Chinook fry have been observed primarily resting on the bottom during night snorkel surveys in Lake Washington (Tabor, pers. comm., 11 July 2000). Reimers (1971) observed that juvenile chinook delayed downstream migration until the darkest part of the night in the Sixes River, Oregon, a result similar to the delay in sockeye migration with increasing light observed by Tabor et al. (1998) in the Cedar River. Key et al. (1994b) found that few juvenile fall chinook were caught during night sampling (relative to diurnal catches) in McNary Reservoir, and proposed that the fish were inactive during night seining. Studies of whether or not chinook or coho juveniles exhibit nocturnal inactivity in Lake Washington were not located. It is possible that the artificial ambient lighting regime in the urbanized basin of Lake Washington may produce uncharacteristic behavior in both juvenile salmonids and their predators. Until more information is available, one should not assume that lighting of overwater or shoreline structures does not affect predator-prey interactions.

Studies from a marine system indicated that wharf lighting could attract juvenile chum and pink salmon (Salo et al. 1977; Prinslow et al. 1979; Prinslow et al. 1980). Significant predation on juvenile chum salmon was not observed in the area of a lighted wharf in Hood Canal, Washington (Prinslow et al. 1980). Whether the behavior of juvenile chinook and coho salmon and their predators in freshwater would behave similarly to chum and pink salmon and their predators in the marine environment remains uncertain and should not be assumed.