The efficacy of fish habitat conservation in land planning processes in Alaska is often constrained by the extent of current knowledge of fish distributions and habitat use. In response to requests for information from land and salmon resource management stakeholders regarding Auke Lake sockeye salmon (Oncorhynchus nerka) status and life history, we examined the prespawning movements and spawning distribution of adult sockeye salmon to provide ecological information needed for Auke Lake watershed management. We used radiotelemetry to track the movements of 80 fish in the Auke Lake watershed during 2012. The prespawning distribution of the fish was not random, indicating five spatially and temporally distinct high-use staging areas within the lake. The Auke Lake sockeye salmon population was dominated by stream-spawning fish (98.5%), with minimal lakeshore spawning (1.5%) observed in association with a small intermittent tributary of the lake. The prespawning distribution patterns identified in this study corroborate observations from Auke Lake 20 y ago, indicating consistent habitat use patterns by sockeye salmon in the lake. Telemetry data also indicate 12% of sockeye salmon expired in Auke Lake without spawning and that 14% of stream-spawning fish were preyed upon by black bear (Usrus americanus). The prespawning and spawning behavior by Auke Lake sockeye salmon suggest that discrete lake staging areas and stream spawning beds are important candidate habitats for protection during the land planning process for shoreline development surrounding the lake.

Populations of sockeye salmon (Oncorhynchus nerka) throughout Southeast Alaska provide locally important subsistence and personal-use fisheries, and contribute to mixed-stock commercial fisheries (Conitz and Cartwright 2005; Heinl et al. 2009; Brunette and Piston 2011; Bednarski et al. 2012). Although many watersheds in Alaska remain relatively isolated from development, others are becoming increasingly urbanized (e.g., Gerken and Sethi 2011). Situated in the City Borough of Juneau, Auke Lake and its saltwater outlet, Auke Bay, support Pacific salmon returns within an urbanizing environment. Auke Bay and the Auke Lake watershed are of significant cultural importance to the Tlingit Áak’w K wáan (e.g., Thornton 2012), and residents of Juneau utilize these systems for fishing, boating, and scientific research, among other uses. Auke Lake sockeye salmon have shown a decreasing trend in abundance since records began in 1963 and have intermittently been the focus of enhancement efforts (Lum and Taylor 2006). Terminal harvest is currently prohibited; however, a proportion of the population is likely harvested in marine mixed-stock commercial gillnet fisheries. As one of only three populations of sockeye salmon in the Juneau vicinity, and the numerically dominant species of Pacific salmon spawning in the upper watershed, Auke Lake sockeye salmon are socially and ecologically important.

As the population of Juneau increases, land development pressure surrounding Auke Lake has continued. At present, 50% of the Auke Lake shoreline is developed (Lum and Taylor 2006) with intensive land uses, such as roads, residential housing, government offices, and the University of Alaska Southeast campus. A walking trail, which has maintained relatively natural habitat, encompasses 30% of the shore; however, natural drainage has been altered and increased human activity is focused at several discrete locations. The remaining 20% of the shore is still in natural condition, but is fragmented into two areas. The lake is managed for multiple uses including recreation with motorized watercraft. Land clearing, storm-water discharge, and road building are considered the factors most likely to contribute to future habitat degradation within the watershed (Juneau Watershed Partnership 2009a, 2009b). In response to requests for information from land and salmon resource management stakeholders regarding Auke Lake sockeye salmon status and life history, we examined the prespawning movements and spawning activity of sockeye salmon within the lake to provide ecological information needed to better manage development in the watershed.

Migration of sockeye salmon into the Auke Lake watershed typically begins in late June and continues through mid-August; however, migration timing does not correspond closely with spawning and many fish reside in the lake for several weeks before spawning (see below). Although extended periods of lake residence prior to spawning are not uncommon in sockeye salmon populations (e.g., Hodgson and Quinn 2002), relatively few studies have documented behavior and habitat use of adult fish during this final phase of their spawning migration. Burger et al. (1995) reported near-shore circling of sockeye salmon along a regular route in Lake Tustamena on the Kenai Peninsula of Southcentral Alaska. Young and Woody (2007a) observed no specific pattern of shoreline migration by sockeye salmon in Lake Clark, Alaska, but documented fish traversing significantly greater distances within the lake than necessary to reach their final spawning location. Working in British Columbia, Roscoe et al. (2010) reported that migration behavior of sockeye salmon differed substantially between two similarly sized interconnected lakes; fish travelled much faster and were less likely to circle and hold in Seton Lake compared to Anderson Lake. Previous studies in Auke Lake by Bucaria (1968) and Nelson (1993) suggested a significant component of the sockeye salmon population (28–48%) may use lakeshore habitat for spawning and indicated that some areas of the lake may be more important for staging fish; however, land development proposals with potential to individually and cumulatively impact Auke Lake shoreline have highlighted significant uncertainty about specific habitat use and distribution of adult sockeye salmon within the lake.

Sockeye salmon can exploit a diverse range of spawning habitats, including lakeshores (Margolis et al. 1995). Lakeshore spawning is not uncommon in sockeye salmon populations throughout Southeast Alaska, with some watersheds (e.g., Kanalku, Sitkoh, and Hetta lakes) dominated by lakeshore-spawning fish (Conitz and Cartwright 2005; Conitz et al. 2007). However, due to the difficulties of working in lakes where visibility is often poor and sockeye salmon spawning depths may exceed 12 m (e.g., Kerns and Donaldson 1968; Olsen 1968), the extent of lakeshore spawning, and thus the importance of lakeshore habitat, is often not well defined. The substrate of Auke Lake is dominated by extensive deposits of fine sediment and decayed vegetative matter. Such substrate conditions are generally not considered suitable spawning habitat for salmonids; however, numerous studies (e.g., Eiler et al. 1992; Young and Woody 2007b) have shown that spawning sockeye salmon can successfully utilize areas such as glacial rivers that were traditionally perceived to be too high in fine sediment to be productive.

In this study, we used radiotelemetry and habitat surveys to provide fine-scale information on lake use and spawning ecology of Auke Lake sockeye salmon to facilitate informed decision making for land planning around Auke Lake. We determined the movements, distribution, and spawning behavior of adult sockeye salmon returning to spawn in Auke Lake and assessed the relative proportions of stream and lakeshore spawning. Specific objectives of this study were as follows: 1) examine habitat use behavior of returning adult sockeye salmon in Auke Lake and identify high-use staging areas; 2) identify whether both stream-spawning and lake-spawning life histories are present in Auke Lake, and determine the relative proportion of sockeye salmon that spawn in Auke Lake vs. Auke Lake inlet streams; and 3) examine spawning phenology of returning sockeye salmon to identify whether different spawning life histories are segregated by run timing. In addition, we provide estimates of prespawn mortality and in-stream bear predation rates.

Study site

Auke Lake (58. °23′17″N 13°37′48″W) is located 16 km northwest of downtown Juneau in Southeast Alaska (Figure 1), with a surface area of 71 ha and a maximum depth of 35 m. The lake has two main inlet streams, Lake Creek and Lake Two Creek, and a single outlet stream, Auke Creek, which flows 0.5 km to salt water at Auke Bay. Lake Creek is the largest of the Auke Lake inlet streams. The surrounding watershed encompasses an area of approximately 1,010 ha, much of which is owned by local, state, and federal government entities. Private residences and the University of Alaska Southeast line the south, west, and north Auke Lake shores. A National Marine Fisheries Service weir located 0.4 km downstream of Auke Lake has monitored fish populations since 1963. Currently, counts of all sockeye salmon migrating into the watershed are recorded.

Figure 1.

Map of study area, showing inlet and outlet streams, fish weir, fixed telemetry locations, and observed locations of radio-tagged sockeye salmon (Oncorhynchus nerka), 2012.

Figure 1.

Map of study area, showing inlet and outlet streams, fish weir, fixed telemetry locations, and observed locations of radio-tagged sockeye salmon (Oncorhynchus nerka), 2012.

Radio-tagging

We sampled adult sockeye salmon ≥400 mm in length and in good condition daily at a fish trap installed at the weir on Auke Creek. We visually assessed all fish that passed the weir for sex. We randomly selected fish from the trap cage and tagged them with a radio transmitter and spaghetti tag for a visual mark. We measured sampled fish for length (mideye to caudal fork) and collected scales for subsequent ageing using protocols following Mosher (1968) and Jearld (1983), which report the number of winters a salmon spent in freshwater and the ocean separated by a decimal point (e.g., 1.2 indicates one winter in freshwater, two winters in ocean). We handled fish with neoprene gloves and supported them underwater with a padded cradle throughout data collection. We inserted radio transmitters into the stomach cavity of fish using a plastic applicator with smooth edges according to the methods described by Eiler et al. (1992); we applied spaghetti tags using a Betadine-sterilized needle applicator. Radio transmitters were model F1835b from Advanced Telemetry Systems (Isanti, MN) and weighed 16 g. Transmitters operated on 8 frequencies with 10 codes per frequency (as described by Kuechle and Kuechle 2012), and we distributed tags on the same frequency among tagging events to help minimize “signal collisions.” We equipped transmitters with a mortality sensor (as described by Eiler 1990) triggered by a nonmovement period of 12 h; transmitters had an expected battery life of 96 d. We deployed 80 radio transmitters throughout the return, representing 5% of the total sockeye salmon return into Auke Lake in 2012. To ensure sampling across both early- and late-run sockeye salmon, the tagging schedule allocated transmitters equally across four strata from June 20, 2012, to August 10, 2012 (Table 1). Fish began passing the weir on June 20 and ceased on August 21; however, only 10 sockeye salmon entered the system after the final tagging event in stratum 4 on August 10. Tag releases began with the first fish entering the fish trap during a given stratum, until the stratum quota of tags were released; we deployed excess or recovered tags from a given strata in the subsequent stratum. When possible, additional fish were sampled at the weir for scales and lengths in order to assess age, sex, and length proportions for both the tagged and nontagged populations.

Table 1.

Tagging dates, numbers of fish tagged, and assigned fates of radio-tagged sockeye salmon (Oncorhynchus nerka) in the Auke Lake watershed, 2012.

Tagging dates, numbers of fish tagged, and assigned fates of radio-tagged sockeye salmon (Oncorhynchus nerka) in the Auke Lake watershed, 2012.
Tagging dates, numbers of fish tagged, and assigned fates of radio-tagged sockeye salmon (Oncorhynchus nerka) in the Auke Lake watershed, 2012.

Transmitter tracking

We tracked tagged fish with a combination of mobile surveys from skiffs and on foot along the lakeshore and streams, and from two receiving stations installed on Lake Creek approximately 50 m and 100 m upstream from the confluence of Auke Lake. Mobile tracking used Advanced Telemetry Systems R4520c data-logging receivers and either a three-element Yagi antenna (boat) or an H antenna (foot). We conducted tracking surveys at least every 3 d from the approximate start of the run on June 23. Tracking ceased on September 1, by which time we had assigned all tagged individuals a fate; however, we continued observations of nontagged fish to the end of spawning activity on September 17. In addition, we conducted three spawning stream surveys in the two primary inlet streams (weeks beginning August 6, 13, and 27), locating radio-tagged fish with mobile tracking gear and visually counting both the tagged and nontagged spawners. We recorded the positions of the radio-tagged fish with a 62stc global position system unit. The activity—swimming, holding, or spawning—of the fish that we could visually was also recorded. The receiving stations automatically recorded movement of radio-tagged fish into and out of Lake Creek, and provided more detailed information on stream entry and timing than the mobile surveys. These stations were equipped with Advanced Telemetry Systems R4500c receivers that continually scanned the transmitter frequencies at 4-s intervals and a four-element Yagi antenna; they were powered with a 12-V deep-cycle battery.

We assigned all tagged fish one of three possible fates: successfully tracked to spawning location, prespawning mortality, and unknown (Table 1). We identified sockeye salmon spawning locations as areas in which we located tagged fish and observed them engaged in spawning behaviors (e.g., redd construction, male–male competition in vicinity of redd, oviposition, redd defense). We identified potential lakeshore spawning areas that could not be visually confirmed as areas where tagged fish remained for 1 wk or more without activating the radio-tag mortality sensor. These areas were subsequently assessed by scuba-based surveys (see below) for spawning activity and the presence of suitable spawning habitat. Prespawn mortalities represented located fish not exhibiting spawning activity or inhabiting a spawning area, and were identified by recovery of a tag from a carcass, or continuous transmission of the mortality signal for three or more days without moving. Radio-tag regurgitation can be difficult to detect (Liedtke and Wargo Rub 2012); however, regurgitation often occurs soon after tagging and thus fish that have been tracked for several days are less likely to regurgitate tags and may be more likely to represent mortalities. We considered transmitter regurgitations as submerged tags recovered without a carcass; however, no such instances were directly observed during the study.

Lake habitat surveys

In order to assess whether spawning activity was occurring in Auke Lake and to assess the suitability of lake habitat for spawning, we conducted scuba-based lake surveys in four fish aggregation sites (Figure 1; Figure S1, Supplemental Material). To increase the likelihood of observing signs of spawning activity, such as evidence of redds, we conducted scuba-based lake surveys after the period of peak spawning activity in the streams; we conducted surveys between August 28 and August 31. We ran survey transects from the shore radially outward to the lake. Divers began at the lake end of transects and swam toward the shore end visually searching for signs of spawning (spawning pairs, active spawning, or redds). Every 5 m, divers recorded the dominant substrate type, following the Wentworth (1922) grain scale, as bedrock, boulder, cobble, large gravel, medium gravel, fine gravel, silt, or organics, and also recorded water depth and depth of organic matter when present. To asses spawning habitat more broadly within the lake and examine both high fish-use and low fish-use lake habitats, we complemented scuba surveys with benthic grab samples collected every 5 m along 26 transects in areas of the lake that received low observed fish use (Figure S2, Supplemental Material).

Data analysis

All statistical analyses were carried out in the R statistical programming environment (RDCT 2013). Using the fates of successfully tracked sockeye salmon as a sample from the general Auke Lake sockeye salmon population, we estimated the proportions of sockeye salmon utilizing different spawning locations, and the proportion of stream- vs. lake-spawning fish. We tested differences in sex composition over temporal strata using Fisher’s exact test. We tested differences in adult fish length over time within sexes using 1-way analysis of variance (ANOVA; separate ANOVAs for male fish, excluding jacks, and for female fish). We tested differences in run timing (based on tagging date) across groups stratified by spawning location (Lake Creek vs. Lake Two Creek), differences in fish that survived to spawn vs. those that exhibited confirmed prespawning mortality (see above), and differences among fish spawning in the upper, middle, and lower reaches of the dominant spawning location, Lake Creek, using Fisher’s exact tests. We mapped the distribution of located fish through time using Gaussian kernel density smoothing using unbiased cross-validation to select the density estimator bandwidth as implemented in the R base stats package (RDCT 2013). We used Ripley’s K analysis to test for spatial randomness of located fish by week of the study, implemented in the spatstat (Baddeley and Turner 2005) R package and using an isotropic edge correction. We assessed significance of spatial non-randomness using a Monte Carlo envelope procedure comparing Ripley’s K from data simulated under complete spatial randomness to that from observed data (e.g., Bivand et al. 2008).

All radio-tagged fish resumed upstream movements along with nontagged fish and entered Auke Lake. We successfully tracked 85% of tagged fish (n  =  68) to spawning locations; 11% (n  =  9) were designated as prespawning mortality, and 4% (n  =  3) were designated as fate unknown (Table 1). Age, sex, and length information collected on radio-tagged and non–radio-tagged fish indicated that the tagged fish constituted a representative sample from the general Auke Lake sockeye salmon population. We successfully aged scales from 293 sockeye salmon, including 71 of the radio-tagged fish. Of the six age classes observed, 2.2 and 2.3 fish made up the dominant age classes in both tagged and nontagged fish (Table 2). Tagging data indicate a small bias toward male fish; however, the proportion of males and females in the tagged vs. nontagged populations were not appreciably different (Table 3). Finally, the length distributions of tagged vs. nontagged fish were not statistically different (Kolmogorov–Smirnov test, P  =  0.1285; Table S1, Supplemental Material). The assigned fate of each radio-tagged fish, along with age, sex, length, and tagging date information is available in Table S2, Supplemental Material.

Table 2.

Age composition of tagged and nontagged sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek weir, 2012.

Age composition of tagged and nontagged sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek weir, 2012.
Age composition of tagged and nontagged sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek weir, 2012.
Table 3.

Sex composition of sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek Weir, 2012.a

Sex composition of sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek Weir, 2012.a
Sex composition of sockeye salmon (Oncorhynchus nerka) sampled at Auke Creek Weir, 2012.a

Spatial distribution

We tracked radio-tagged sockeye salmon in Auke Lake from June 23 to September 1. Sockeye salmon lake location and movement data presented here is from the 9 wk period June 23 to August 26. Auke Lake habitat use by returning adult salmon was nonrandom (Figure 2; Figure S3, Supplemental Material). Lake use followed a progression of early staging along the southern and eastern shores (Figures 2a–2c), with relatively diffuse distribution of salmon (Figures S3a–S3d, Supplemental Material), followed by concentrated aggregations of fish holding on the northern and northwestern shores close to the inlet spawning streams (Figures 2d–2i; Figures S3e–S3i, Supplemental Material). By week 5 of the study (July 21–July 27), nearly all radio-tagged fish were concentrated along the northern shore of Auke Lake in proximity to inlet spawning streams (Figure 2e; Figure S3e, Supplemental Material). Fish remaining in the lake at the end of the study period in week 9 (August 18-August 26) missed initial access to the main spawning streams and showed a propensity to wander the lake (Figure 2i: see below).

Figure 2.

Gaussian kernel smoothing of sockeye salmon (Oncorhynchus nerka) locations by weekly time strata. Smoothed intensities are scaled to the maximum intensity. As the season progressed, sockeye salmon holding areas shifted from the east shore to the south shore, followed by concentration of salmon about the spawning streams located in the north end of the lake. Red circles indicate fish locations, green points indicate stream inlets and the blue dot indicates the lake outlet at Auke Creek. Weekly dates, 2012: (a) week 1: June 23–June 29; (b) week 2: June 30–July 6; (c) week 3: July 7–July 13; (d) week 4: July 14–July 20; (e) week 5: July 21–July 27; (f) week 6: July 28–August 3; (g) week 7: August 4–August 10; (h) week 8: August 11–August 17; (i) week 9: August 18–August 26.

Figure 2.

Gaussian kernel smoothing of sockeye salmon (Oncorhynchus nerka) locations by weekly time strata. Smoothed intensities are scaled to the maximum intensity. As the season progressed, sockeye salmon holding areas shifted from the east shore to the south shore, followed by concentration of salmon about the spawning streams located in the north end of the lake. Red circles indicate fish locations, green points indicate stream inlets and the blue dot indicates the lake outlet at Auke Creek. Weekly dates, 2012: (a) week 1: June 23–June 29; (b) week 2: June 30–July 6; (c) week 3: July 7–July 13; (d) week 4: July 14–July 20; (e) week 5: July 21–July 27; (f) week 6: July 28–August 3; (g) week 7: August 4–August 10; (h) week 8: August 11–August 17; (i) week 9: August 18–August 26.

Sockeye salmon held in Auke Lake prior to spawning for an average of 29 d, with fish residence time ranging from 5 to 45 d. Fish that entered the lake early in the study during the first release stratum tended to hold longer in the lake prior to spawning and explored more of the lake, particularly in the southern and eastern areas (Table 4; Figure 3a). Subsequent to the first release stratum, the majority of fish followed a movement pattern of entering the lake, heading quickly for the northern shore where the inlet spawning streams were located, and then transiting laterally between holding aggregations (Figures 3b–3d).

Table 4.

Days between entry into Auke Lake and ascension into Lake Creek for sockeye salmon (Oncorhynchus nerka) radio-tagged at Auke Creek weir, 2012.

Days between entry into Auke Lake and ascension into Lake Creek for sockeye salmon (Oncorhynchus nerka) radio-tagged at Auke Creek weir, 2012.
Days between entry into Auke Lake and ascension into Lake Creek for sockeye salmon (Oncorhynchus nerka) radio-tagged at Auke Creek weir, 2012.
Figure 3.

Radio-tagged sockeye salmon (Oncorhynchus nerka) movements (thin gray lines) between successive locations in Auke Lake (thick outline), 2012. (a) Stratum 1: June 20–July 1; (b) stratum 2: July 2–July 13; (c) stratum 3: July 14–July 25; (d) stratum 4: July 26–August 10.

Figure 3.

Radio-tagged sockeye salmon (Oncorhynchus nerka) movements (thin gray lines) between successive locations in Auke Lake (thick outline), 2012. (a) Stratum 1: June 20–July 1; (b) stratum 2: July 2–July 13; (c) stratum 3: July 14–July 25; (d) stratum 4: July 26–August 10.

Aggregating fish distribution data across the study, we visually identified four discrete high-use sites within Auke Lake during the commencement of spawning activity (Figure 1; Figure S1, Supplemental Material). Two high-use sites were located at the mouths of Lake Creek and Lake Two Creek, whereas two were not associated with streams (Figure 1; Figure S1, Supplemental Material). Scuba surveys did not, however, identify spawning activity at any of these high-use sites and indicated that three of the sites (sites 2, 3, and 4, Figure 1; Figure S1 and Table S3, Supplemental Material) presented unsuitable sockeye salmon spawning habitat. Only high-use site 1 contained a limited amount of poor quality spawning habitat.

Spawning activity and phenology

Using the fates of successfully tracked sockeye salmon as a sample from the general population, 98.5% (n  =  67) of sockeye salmon entering Auke Lake watershed in 2012 spawned in stream environments, whereas only 1.5% (n  =  1) spawned in the lake environment (Table 5). Of the stream spawners, 88.1% spawned in Lake Creek and 9.0% spawned in Lake Two Creek. Only two tagged fish that spawned in stream environments utilized habitat outside of the two primary inlet creeks: one spawned in an unnamed ephemeral tributary in the southeastern quadrant of the lake and one female constructed a redd in Auke Creek (Figure 1). Four nontagged sockeye salmon were also observed spawning in the unnamed tributary, whereas no other sockeye salmon spawned in Auke Creek.

Table 5.

Spawning location proportions for radio-tagged sockeye salmon (Oncorhynchus nerka) successfully fated to any spawning location in the Auke Lake watershed, 2012.

Spawning location proportions for radio-tagged sockeye salmon (Oncorhynchus nerka) successfully fated to any spawning location in the Auke Lake watershed, 2012.
Spawning location proportions for radio-tagged sockeye salmon (Oncorhynchus nerka) successfully fated to any spawning location in the Auke Lake watershed, 2012.

A single radio-tagged fish was confirmed spawning in Auke Lake; mobile tracking during the study observed four other nontagged fish spawning at this lake site. The single confirmed lake spawning site was immediately adjacent to the ephemeral stream that appeared in the southeastern quadrant of the lake (Figure 1).

The receiving stations located at the mouth of Lake Creek demonstrated that entry into stream environments to spawn was mediated by water connectivity between the lake and stream in 2012. Fish ascended Lake Creek in two pulses, beginning with a high flow event on July 31 in which connectivity between the stream and lake persisted for 17 d, accounting for passage of 80% of tagged fish spawning in Lake Creek (n  =  47); for 11 d prior to this no surface water connectivity existed between Lake Creek and Auke Lake. After a week-long period of no surface water connectivity between the stream and lake, the remaining 20% of Lake Creek spawners (n  =  12) ascended the stream after a heavy rain event (August 22).

Consistent with sockeye salmon life history characteristics observed in other Alaskan and North Pacific systems (e.g., Morbey 2000; Quinn 2005), there was a difference in the sex composition of fish entering Auke Creek over time (Fisher’s exact test, P ≪ 0.001), with an increasing proportion of female fish as the season progressed (53, 60, 64, and 75% proportion female across temporal strata 1–4, respectively). Furthermore, 1-way ANOVA indicated a statistically significant difference in the length of adult fish entering Auke Creek over time for both females and males, with lengths within sexes decreasing across temporal strata (females: F3,178  =  22.08, P ≪ 0.001; males: F3,123  =  14.44, P ≪ 0.001). Females averaged 546.5 mm, 524.4 mm, 510.5 mm, and 501.4 mm in strata 1–4, respectively. Males averaged 568.7 mm, 543.1 mm, 519.5 mm, and 515.3 mm in strata 1-4, respectively. Decreasing length in late-returning compared to early-returning fish is frequently observed within Pacific salmon populations (Quinn 2005).

No difference in run timing was indicated between fish that ultimately spawned in Lake Creek vs. Lake Two Creek (Fisher’s exact test, P  =  0.64). Due to low counts of tagged fish observed spawning in other streams, we were unable to test for differences in run timing; however, stream surveys indicated that peak spawning activity for Lake Creek and Lake Two Creek occurred over the week of August 13 to August 19, whereas spawning in the unnamed ephemeral creek and adjacent lake spawning habitat occurred 3 wk later, from September 3 to September 9.

Early entrants to Auke Lake tended to spawn higher up in Lake Creek, even though water levels in the stream appeared sufficient for later-entering fish to ascend as far. Dividing up Lake Creek into three equal length reaches (378 m/reach), we found a significant difference in run timing among fish fated to upper, middle, and lower spawning areas (P  =  0.025), with fish entering Auke Lake during tagging strata 1 and 2 preferring middle and upper reaches of Lake Creek and fish entering during tagging strata 3 and 4 primarily utilizing the lower reach of Lake Creek.

Mortality

Using the radio-tagged fish designated as mortalities as a sample from the general population, 12% (9 of 77; 95% CI: [6%, 20%]) of sockeye salmon died in Auke Lake in 2012. We did not observe these fish spawning in any lake or stream habitat during extensive radio tracking and assume that they represent prespawning mortality. We confirmed this assessment for five fish recovered as mortalities. We tracked the four remaining fish in the lake for 8–46 d prior to transmitting a mortality signal, suggesting that these fish were mortalities rather than regurgitated tags. Although we found no significant difference in run timing between fish that survived to spawn and prespawning mortalities (Fisher’s exact test, P  =  0.08), later-entering fish appeared to make up a larger proportion of the prespawning mortalities. Foot surveys along the two primary spawning streams located black bear (Ursus americanus) predation of both radio-tagged and nontagged salmon. Using tagged fish as a sample from the general population, 14% (9 of 65 tagged creek spawners; 95% CI: [7%, 24%]) of stream-spawning sockeye salmon returning to Auke Lake were preyed upon by bears.

Based on the radio-tagged sample, the Auke Lake sockeye salmon return in 2012 was dominated by fish spawning in stream environments (98.5%), whereas only a minor component (1.5%) of spawners used lakeshore habitat. This was consistent with work in Auke Lake by Bucaria (1968), who estimated 99% of spawners used stream habitat and 1% used lakeshore habitat. Lakeshore spawning is not uncommon throughout Southeast Alaska; however, the degree to which it occurs is variable, with some watersheds (e.g., Kanalku, Sitkoh, and Hetta lakes) exhibiting large proportions (>50%) of lakeshore-spawning fish (Conitz and Cartwright 2005; Conitz et al. 2007), whereas others (e.g., McDonald Lake) have low proportions (∼10%; Heinl et al. 2009), or have no documented lakeshore spawning (e.g., Klawock and Eek Lakes; Conitz et al. 2007; Conitz 2008). Sockeye salmon can use a diverse range of spawning habitats and likely exploit lakeshore spawning habitat to the extent it is available in any given watershed. Scuba surveys and benthic grab samples in Auke Lake indicated limited availability of suitable spawning substrate throughout most of the lakeshore, with the benthos dominated by glacial and organic silt. Although the proportion of lake spawners may vary annually as a function of spawner abundance or access to inlet streams, it is unlikely the lake has sufficient habitat capacity to support a significant increase in the proportion of spawners successfully using the lakeshore.

Lakeshore-spawning sockeye salmon typically select areas with groundwater influence and frequently spawn later than tributary-spawning fish within the same watershed (Burgner 1991). The limited lakeshore spawning that we observed in Auke Lake was primarily associated with a small intermittent tributary with fish likely attracted to groundwater or hyporheic influence at the lake–tributary interface. Based upon earlier reports, sockeye salmon may have been using this site for at least 49 y (Bucaria 1968). Similarly, Heinl et al. (2009), working in McDonald Lake, Southeast Alaska, observed that all lakeshore spawning by sockeye salmon was associated with very small intermittent tributaries. In 2012, the limited spawning activity observed in the lakeshore habitat of Auke Lake occurred 3 wk later than peak spawning activity in the two main inlet streams. Heinl et al. (2009) documented that lakeshore spawning was initiated 4 wk later than tributary spawning in McDonald Lake. Temporal and spatial separation of spawning groups can promote genetic differentiation and population substructure, with tributary- and lake-spawning sockeye salmon often comprising discrete spawning populations (e.g., Varnavskaya et al. 1994). Sockeye salmon spawning in the small intermittent tributary and associated lakeshore habitat may consist of an at least partially isolated spawning group and represent life history diversity within the Auke Lake population.

Among the stream-environment spawning sockeye salmon in Auke Lake, 88.1% of fish utilized Lake Creek. Receiving stations and manual tracking showed sockeye salmon access to Lake Creek is synchronized with stream flow. In 2012, fish entered the stream in two pulses, separated by a week with no stream access. Nelson (1993) observed the same pattern in Auke Lake in 1992, whereas in 1991 stream flow was sufficient for continuous access throughout the spawning period. In 2012, during the second period of restricted creek access, a few sockeye salmon were observed constructing redds on the gravel delta formed by Lake Creek. This low-flow period followed the high flows that facilitated movement of most fish (80% of tagged fish) into Lake Creek. Sockeye salmon spawning on the delta were likely homing to Lake Creek, but spawned on the delta by default when access to the creek was restricted. Opportunistically spawning on the gravel delta of Lake Creek, and to a lesser extent, Lake Two Creek, may allow sockeye salmon to spawn when stream access is restricted by low flow, or when spawner abundance is sufficiently high to limit available stream habitat. The wetted area of the Lake Creek delta may offer high-quality spawning habitat, in terms of substrate composition and hyporheic flow, during portions of the summer; however, depending on flow, redds are likely susceptible to desiccation during parts of the summer and winter. Further, in contrast to the lakeshore spawning site associated with the intermittent tributary, the deltas are seasonally highly unstable, likely resulting in low embryo survival. Of the six redds constructed on the delta in 2012, we suspected that five were destroyed by a late September flood event.

Sockeye salmon held in Auke Lake prior to spawning for an average of 29 d, with fish residence time ranging from 5 to 45 d. During residency in Auke Lake, habitat use by returning adult salmon was nonrandom (Figure 2; Figure S2, Supplemental Material). The distribution of adult sockeye salmon observed in Auke Lake in 2012 was remarkably consistent with observations by Nelson (1993) in 1991 and 1992. Staging along the southern and eastern shore habitats early in the study period was documented in both studies. Our results indicate these areas were primarily used by early-returning fish in stratum 1 (June 20 to July 1; Table 1). Sockeye salmon in subsequent return strata, particularly strata 3 and 4, moved rapidly to the northern shore of the lake where the inlet spawning streams were located, possibly reflecting more advanced sexual maturation on entering the lake (Figure 3). Nelson (1993) also documented three of the four discrete high-use sites (sites 1, 2, and 4) identified on the northern and western shores, as well as a single site on the southwestern shore not observed in 2012. During this study sockeye salmon distribution was generally more diffuse in the first half of the study period, followed by concentrated aggregations of fish holding on the northern and northwestern shores close to the inlet spawning streams. Using acoustic telemetry, Newell (2005) observed a similar pattern of distribution for sockeye salmon in Lake Washington, Washington. During their first 4 to 6 wk in the lake, fish moved throughout the lake, whereas later in the summer and fall most fish were concentrated to the west and south of Mercer Island, presumably staging to ascend their spawning tributary.

Nelson (1993) suggested that, based on opportunistic scuba surveys, high-use sites observed in 1991–1992 might represent lake-based spawning aggregations. However, the systematic benthic scuba sampling design employed in 2012 did not identify spawning activity in the lake at any of the four fish aggregation sites and confirmed that three sites would not support spawning activity, with marginal spawning habitat at the remaining site. This provides strong evidence that habitat criteria important for staging behavior likely structured the observed distribution of sockeye salmon in Auke Lake, and that the fish were not selecting aggregation areas based on their suitability for spawning. Fish frequently stage at stream mouths prior to entering them to spawn, and creek proximity was a strong factor influencing the selection of sites at the mouths of Lake and Lake Two creeks by staging sockeye salmon; however, factors influencing selection of the two high-use sites not associated with streams remain unclear. The balance between energy expenditure and reproductive maturation during the final stage of migration is an important factor that may induce sockeye salmon to seek out specific thermal conditions within natal lakes (Roscoe et al. 2010). For example, Newell and Quinn (2005) reported that sockeye salmon in Lake Washington primarily occupied the hypolimnion between 9 and 11°C, with fish rarely using cooler or warmer water. Use of this temperature range may represent an optimal balance between metabolic energy expenditure and reproductive maturation (Newell and Quinn 2005). Although water temperature data were not available to test hypotheses about staging areas as thermal refugia, it is possible such a dynamic may be driving Auke Lake sockeye salmon staging behavior. Regardless of mechanism, continued use of sites over time as observed in this study suggests Auke Lake is not a homogenous habitat and sockeye salmon are preferentially selecting areas for staging.

Field sampling for this project took place during a single year. Weather conditions were cooler and significantly wetter than average, prior to and during the early part of the study period (May 1–July 29); however, stream flow patterns that we observed in the two main inlet streams throughout the study period closely resembled those described in other studies (Bucaria 1968; Nelson 1993). Furthermore, fewer sockeye salmon (N  =  1,565) passed the Auke Creek weir in 2012, compared to the 10-y average (N  =  2,655). We are confident the results from 2012 are representative of Auke Lake sockeye salmon behavior because of the consistency in fish behavior that we observed with similar studies conducted across a range of fish abundance and flow conditions by Bucaria (1968) sampling in 1963–1964 and Nelson (1993) sampling in 1991–1992.

This study has shown that, in addition to providing important rearing habitat for juvenile sockeye salmon, the returning adults concentrate use in portions of Auke Lake and remain in the lake for an extended period prior to spawning. Areas of particular importance for staging Auke Lake sockeye salmon appear to be the southern bay and the four sites identified on the northern and western shores. We did not evaluate specific habitat criteria for selection of staging areas by sockeye salmon in this study. Future investigation into habitat factors, such as water temperature, dissolved oxygen, and lake flow patterns that are important in structuring the staging distribution of sockeye salmon will be useful in assessing potential impacts of development—such as simplification of the shoreline through habitat conversion or displacement of staging salmon into suboptimal areas as a result of high motorized boat activity—on the sockeye salmon population in Auke Lake, as well as provide information on spawning sockeye salmon lake use in other comparable systems.

Spawning behavior observations during this study highlighted the significance of the two main inlet streams, particularly Lake Creek, to Auke Lake sockeye salmon. For example, in 2012, 87% of spawning-fated radio-tagged sockeye salmon spawned in the 1.1 km of Lake Creek immediately upstream of Auke Lake. Sustainability of Auke Lake sockeye salmon is dependent on spatially limited spawning habitat, which exposes the population to significant risk from natural events such as flooding and drought, but also anthropogenic impacts. In 2012, 67% of all tagged fish and 81% of later-run (strata 3 and 4) tagged fish spawned in the first 0.5 km of Lake Creek, which is channelized, surrounded by residential subdivisions, and intersected by a road. Development activities that reduce flow or degrade stream and riparian functions in this reach could significantly impact the population and would likely disproportionately impact late-run fish. The potential for adverse effects on a spatially concentrated area of spawning activity highlights the importance of adequately protecting hydrological, stream channel, and riparian functions of the inlet streams from individual and cumulative anthropogenic impacts, as well as restoring degraded habitat where feasible.

Evidence from this study and that of Heinl et al. (2009) suggests that even very small intermittent tributaries of lakes in Southeast Alaska, and possibly elsewhere, have an important role in creating suitable lakeshore spawning habitat for sockeye salmon. These habitats may promote life history diversity and resilience of sockeye salmon populations. The role of intermittent streams within aquatic ecosystem networks has been poorly studied and undervalued (Larned et al. 2010); however, their significance to fish populations, including salmonids, is becoming more widely documented (e.g., Erman et al. 1976; Maslin et al. 1998; Wigington et al. 2006). Intermittent streams have been shown to be more sensitive to degradation than larger perennial streams; however, at present they receive less regulatory protection (see McDonough et al. 2011). Land managers and regulatory agencies should consider intermittent streams as potentially important salmonid habitat for protection during land planning processes.

Land development processes in urbanizing areas like Auke Lake can lead to cumulative impacts from numerous relatively small projects that degrade habitats or displace sockeye salmon from preferred habitats. Recent and ongoing development projects around Auke Lake include construction of a boat launch, walking trails, office buildings, and road improvements. In some cases, these development projects have been sited near high-use areas for sockeye salmon staging or near the intermittent stream where lake spawning was observed. Although shoreline development and habitat integrity are not mutually exclusive, future studies to understand the cumulative impacts of development on sockeye salmon life history, including spawning success and juvenile survival, would provide better understanding of how to best manage Auke Lake and the surrounding watershed to ensure long-term sustainability of an important urban sockeye salmon run.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Figure S1. Sockeye salmon (Oncorhynchus nerka) aggregation sites during spawning activity and scuba transect locations, Auke Lake, 2012. Dots represent sockeye locations across the study period from June 23, 2012, to September 1, 2012.

Found at DOI: 10.3996/112014-JFWM-083.s1 (97 KB DOCX).

Figure S2. Location of benthic grab sample transects, Auke Lake 2012.

Found at DOI: 10.3996/112014-JFWM-083.s2 (2472 KB DOCX).

Figure S3. Ripley’s K plots to test for complete spatial randomness in Auke Lake sockeye salmon (Oncorhynchus nerka) locations by week-long strata, 2012. Weekly dates, 2012: 1: June 23–June 29; 2: June 30–July 6; 3: July 7–July 13; 4: July 14–July 20; 5: July 21–July 27; 6: July 28–August 3; 7: August 4–August 10; 8: August 11–August 17; 9: August 18–August 26. Ripley’s K is an index measuring the expected number of points about any given point in a radius, r. Clustered points have K values greater than that expected under complete spatial randomness. Gray shaded envelope polygons represent Monte Carlo simulations under a theoretical complete spatial random process for the data; observed Ripley’s K lines above the theoretical expected values and envelope polygons are evidence of clustered locations. As the season progresses, salmon cluster about spawning streams and holding areas on the north shore and off a point on the west shore (see main text and Figure 2), as indicated by increasing distance between the observed and theoretical K values.

Found at DOI: 10.3996/112014-JFWM-083.s3 (231 KB DOCX).

Table S1. Length frequency distribution of nontagged sockeye salmon (Oncorhynchus nerka) and sockeye salmon implanted with radio transmitters, Auke Lake, 2012.

Found at DOI: 10.3996/112014-JFWM-083.s4 (15 KB DOCX).

Table S2. Assigned fates, tagging dates, and age, sex, length information for each radio-tagged sockeye salmon (Oncorhynchus nerka), Auke Lake, 2012.

Found at DOI: 10.3996/112014-JFWM-083.s5 (29 KB DOCX).

Table S3. Lakeshore habitat scuba surveys, Auke Lake, 2012.

Found at DOI: 10.3996/112014-JFWM-083.s6 (14 KB DOCX).

Reference S1. Bednarski J, Harris DK, Monagle K, Heinl SC, Kelley MS. 2012. Northern Chatham Strait sockeye salmon: updated stock status, fishery management, and subsistence fisheries. Alaska Department of Fish and Game, Division of Commercial Fisheries, Regional Information Report IJ12-14, Douglas, Alaska.

Found at DOI: 10.3996/112014-JFWM-083.s7; also available at http://www.adfg.alaska.gov/FedAidpdfs/RIR.1J.2012.14.pdf (1516 KB PDF).

Reference S2. Brunette MT, Piston AW. 2011. Hugh Smith Lake sockeye salmon studies, 2010. Alaska Department of Fish and Game Fishery Data Series Number 11-32, Anchorage.

Found at DOI: 10.3996/112014-JFWM-083.s8; also available at http://www.sf.adfg.state.ak.us/FedAidPDFs/FDS11-32.pdf (2010 KB PDF).

Reference S3. Bucaria GP. 1968. Growth of juvenile sockeye salmon, Oncorhynchus nerka (Walbaum), in Auke Lake, Alaska. Master’s thesis. Oregon State University, Corvallis.

Found at DOI: 10.3996/112014-JFWM-083.s9; also available at http://ir.library.oregonstate.edu/xmlui/handle/1957/22243 (1752 KB PDF).

Reference S4. Conitz JM. 2008. Klawock Lake subsistence sockeye salmon project 2006 annual report and 2004-2006 summary. Alaska Department of Fish and Game Fishery Data Series Number 08-48, Anchorage.

Found at DOI: 10.3996/112014-JFWM-083.s10; also available at http://www.sf.adfg.state.ak.us/FedAidPDFs/fds08-48.pdf (699 KB PDF).

Reference S5. Conitz JM, Cartwright MA. 2005. Kanalku, Sitkoh, and Kook Lakes subsistence sockeye salmon project: 2003 Annual Report and 2001–2003 Final Report. Alaska Department of Fish and Game Fishery Data Series Number 05-57, Anchorage.

Found at DOI: 10.3996/112014-JFWM-083.s12; also available at http://www.adfg.alaska.gov/fedaidpdfs/fds05-57.pdf (513 KB PDF).

Reference S6. Conitz JM, Stahl JM, Bale RW, Cartwright MA. 2007. Hetta and Eek Lakes subsistence sockeye salmon project: 2004 Annual Report. U.S. Fish and Wildlife Service Office of Subsistence Management, Fisheries Resource Monitoring Program (Study 04-606), Anchorage, Alaska.

Found at DOI: 10.3996/112014-JFWM-083.s11; also available at http://www.adfg.alaska.gov/fedaidpdfs/fds07-19.pdf (1043 KB PDF).

Reference S7. Gerken J, Sethi SA. 2011. Juvenile coho salmon migration and habitat use in Meadow Creek, South-central Alaska, 2011. U.S. Fish and Wildlife Service Data Series Report 2013-1, Anchorage, Alaska.

Found at DOI: 10.3996/112014-JFWM-083.s13; also available at http://www.fws.gov/alaska/fisheries/fish/Data_Series/d_2013_ 1.pdf (5793 KB PDF).

Reference S8. Heinl SC, Eggers DM, Piston AW. 2009. Sockeye salmon mark-recapture and radio telemetry studies at McDonald Lake in 2007. Alaska Department of Fish and Game Fishery Data Series Number 09-42, Anchorage.

Found at DOI: 10.3996/112014-JFWM-083.s14; also available at http://www.sf.adfg.state.ak.us/FedAidPDFs/FDS09-42.pdf (1497 KB PDF).

Reference S9. Juneau Watershed Partnership. 2009a. Auke Lake Watershed Assessment.

Found at DOI: 10.3996/112014-JFWM-083.s15; also available at http://www.seakfhp.org/wp-content/uploads/2013/03/Auke-Lake-Assessment-2009.pdf (15.8 MB PDF).

Reference S10. Juneau Watershed Partnership. 2009b. Auke Lake Watershed Action Plan.

Found at DOI: 10.3996/112014-JFWM-083.s16; also available at http://www.seakfhp.org/wp-content/uploads/2013/03/Auke-Lake-Action-Plan-2009.pdf (895 KB PDF).

Reference S11. Lum J, Taylor S. 2006 Dolly Varden and cutthroat trout migrations at Auke Creek in 2003, and abundance of cutthroat trout in Auke Lake, southeast Alaska. Alaska Department of Fish and Game, Divisions of Sport Fish Research and Technical Services, Anchorage.

Found at DOI: 10.3996/112014-JFWM-083.s17; also available at http://www.sf.adfg.state.ak.us/FedAidPDFs/fds06-09.pdf (500 KB PDF).

Reference S12. Maslin PE, McKinney WR, Moore TL. 1998. Intermittent streams as rearing habitat for Sacramento River Chinook salmon. Unpublished report prepared for the U.S. Fish and Wildlife Service under the authority of the Federal Grant and Cooperative Agreement Act of 1977 and the Central Valley Improvement Act.

Found at DOI: 10.3996/112014-JFWM-083.s18; also available at http://www.calwater.ca.gov/Admin_Record/D-022206.pdf (857 KB PDF).

Reference S13. Nelson B. 1993. Movements of radio-tagged sockeye salmon in Auke Lake, southeast Alaska. National Marine Fisheries Service Auke Bay Laboratory. Unpublished Draft Technical Memorandum.

Found at DOI: 10.3996/112014-JFWM-083.s19 (2524 KB PDF).

Reference S14. Newell J. 2005. Migration and movement patterns of adult sockeye salmon (Oncorhynchus nerka) in Lake Washington. Master’s thesis. University of Washington, Seattle.

Found at DOI: 10.3996/112014-JFWM-083.s20; also available at http://weber.s.uw.edu/research/publications/ms_phd/Newell_J_MS_Sp05.pdf (1223 KB PDF).

This project was funded under the Alaska Sustainable Salmon Fund with additional contributions from the Alaska Department of Transportation and Public Facilities.

We thank the following for assistance with field work: J. Parish and B. Wirth (University of Alaska Southeast); B. Heard, P. Malacha, E. Siddon, and S. Vulstek (National Marine Fisheries Service–Alaska Fisheries Science Center); F. Pryor (Alaska Department of Fish and Game); and S. Brockmann, D. Evans, and J. Hudson (U.S. Fish and Wildlife Service). Thanks to N. Stichert (USFWS) for logistical support and critical review of an earlier draft of this manuscript. We thank Bonita Nelson (National Marine Fisheries Service–Alaska Fisheries Science Center) for useful insight into the study area. We thank two anonymous reviewers for comments that improved this manuscript. We are grateful to the University of Alaska Southeast for access to the Auke Lake dock, and the City of Juneau for assistance with aerial imagery.

Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Baddeley
A
,
Turner
R
.
2005
.
Spatstat: an R package for analyzing spatial point patterns
.
Journal of Statistical Software
6
:
1
42
.
Bednarski
J
,
Harris
DK
,
Monagle
K
,
Heinl
SC
,
Kelley
MS
.
2012
.
Northern Chatham Strait sockeye salmon: updated stock status, fishery management, and subsistence fisheries
.
Douglas, Alaska: Alaska Department of Fish and Game, Division of Commercial Fisheries, Regional Information Report IJ12-14 (see Supplemental Material, Reference S1, 10.3996/112014-JFWM-083.S7); also
.
Bivand
RS
,
Pebesma
EJ
,
Gomez-Rubio
V
.
2008
.
Applied spatial data analysis with R
.
New York
:
Springer
.
Brunette
MT
,
Piston
AW
.
2011
.
Hugh Smith Lake sockeye salmon studies, 2010
.
Anchorage: Alaska Department of Fish and Game Fishery Data Series Number 11-32 (see Supplemental Material, Reference S2, 10.3996/112014-JFWM-083.S8); also available: http://www.sf.adfg.state.ak.us/FedAidPDFs/FDS11-32.pdf (October 2013)
.
Bucaria
GP
.
1968
.
Growth of juvenile sockeye salmon, Oncorhynchus nerka (Walbaum), in Auke Lake, Alaska
.
Master’s thesis
.
Corvallis
:
Oregon State University
(see Supplemental Material, Reference S3, 10.3996/112014-JFWM-083.S9); also available: http://ir.library.oregonstate.edu/xmlui/handle/1957/22243 (January 2012)
.
Burger
CV
,
Finn
JE
,
Holland-Bartels
L
.
1995
.
Pattern of shoreline spawning by sockeye salmon in a glacially turbid lake: evidence for population differentiation
.
Transactions of the American Fisheries Society
124
:
1
15
.
Burgner
RL
.
1991
.
Life history of sockeye salmon (Oncorhynchus nerka)
. Pages
1
118
in
Groot
C
,
Margolis
L.
,
editors. 1991
.
Pacific salmon life histories
.
Vancouver, Canada
:
UBC Press
.
Cherry
S
.
1996
.
A comparison of confidence interval methods for habitat use-availability studies
.
Journal of Wildlife Management
60
:
653
658
.
Conitz
JM
.
2008
.
Klawock Lake subsistence sockeye salmon project 2006 annual report and 2004-2006 summary
.
Anchorage: Alaska Department of Fish and Game Fishery Data Series Number 08-48 (see Supplemental Material, Reference S4, 10.3996/112014-JFWM-083.S10); also available: http://www.sf.adfg.state.ak.us/FedAidPDFs/fds08-48.pdf (October 2013)
.
Conitz
JM
,
Cartwright
MA
.
2005
.
Kanalku, Sitkoh, and Kook Lakes subsistence sockeye salmon project: 2003 Annual Report and 2001–2003 Final Report
.
Anchorage: Alaska Department of Fish and Game Fishery Data Series Number 05-57 (see Supplemental Material, Reference S5, 10.3996/112014-JFWM-083.S12); also available: http://www.adfg.alaska.gov/fedaidpdfs/fds05-57.pdf (October 2013)
.
Conitz
JM
,
Stahl
JM
,
Bale
RW
,
Cartwright
MA
.
2007
.
Hetta and Eek Lakes subsistence sockeye salmon project: 2004 Annual Report
.
Anchorage, Alaska: U.S. Fish and Wildlife Service Office of Subsistence Management, Fisheries Resource Monitoring Program (Study 04-606) (see Supplemental Material, Reference S6, 10.3996/112014-JFWM-083.S11); also available: http://www.adfg.alaska.gov/fedaidpdfs/fds07-19.pdf (October 2013)
.
Eiler
JH
.
1990
.
Radio transmitters used to study salmon in glacial rivers
. Pages
364
369
in
Parker
N
,
editor.
Fish-marking techniques
.
Bethesda, Maryland
:
American Fisheries Society Symposium 7
.
Eiler
JH
,
Nelson
BD
,
Bradshaw
RF
.
1992
.
Riverine spawning by sockeye salmon in the Taku River, Alaska and British Columbia
.
Transactions of the American Fisheries Society
121
:
701
708
.
Erman
DC
,
Hawthorne
VM
.
1976
.
The quantitative importance of an intermittent stream in the spawning of rainbow trout
.
Transactions of the American Fisheries Society
105
:
675
681
.
Gerken
J
,
Sethi
SA
.
2011
.
Juvenile coho salmon migration and habitat use in Meadow Creek, South-central Alaska, 2011
.
Anchorage, Alaska: U.S. Fish and Wildlife Service Data Series Report 2013-1 (see Supplemental Material, Reference S7, 10.3996/112014-JFWM-083.S13); also available: http://www.fws.gov/alaska/fisheries/fish/Data_Series/d_2013_1.pdf (November 2013)
.
Heinl
SC
,
Eggers
DM
,
Piston
AW
.
2009
.
Sockeye salmon mark-recapture and radio telemetry studies at McDonald Lake in 2007
.
Anchorage: Alaska Department of Fish and Game Fishery Data Series Number 09-42 (see Supplemental Material, Reference S8, 10.3996/112014-JFWM-083.S14); also available: http://www.sf.adfg.state.ak.us/FedAidPDFs/FDS09-42.pdf (October 2013)
.
Hodgson
S
,
Quinn
TP
.
2002
.
The timing of adult sockeye salmon migration into freshwater: adaptations by populations to prevailing thermal regimes
.
Canadian Journal of Zoology
80
:
542
555
.
Jearld
A
.
1983
.
Age determination
. Pages
301
324
in
Nielsen
LA
,
Johnson
DL
.,
editors.
Fisheries techniques
.
Bethesda, MD
:
American Fisheries Society
.
Juneau Watershed Partnership
.
2009a
.
Auke Lake Watershed Assessment (see Supplemental Material, Reference S9, 10.3996/112014-JFWM-083.S15); also available: http://www.seakfhp.org/wp-content/uploads/2013/03/Auke-Lake-Assessment-2009.pdf (November 2011)
.
Juneau Watershed Partnership
.
2009b
.
Auke Lake Watershed Action Plan (see Supplemental Material, Reference S10, 10.3996/112014-JFWM-083.S15); also available: http://www.seakfhp.org/wp-content/uploads/2013/03/Auke-Lake-Action-Plan-2009.pdf (November 2011)
.
Kerns
OE
Jr
,
Donaldson
JR
.
1968
.
Behavior and distribution of spawning sockeye salmon on island beaches in lliamna Lake, Alaska
.
Journal of the Fisheries Research Board of Canada
25
:
485
494
.
Kuechle
VB
,
Kuechle
PJ
.
2012
.
Radio telemetry in fresh water: the basics
. Pages
91
137
in
Adams
NS
,
Beeman
JW
,
Eiler
JH
,
editors.
Telemetry techniques: a user guide for fisheries research
.
Bethesda, Maryland
:
American Fisheries Society
.
Larned
ST
,
Datry
T
,
Arscott
DB
,
Tockner
K
.
2010
.
Emerging concepts in temporary-river ecology
.
Freshwater Biology
55
:
717
738
.
Liedtke
TL
,
Wargo Rub
AM
.
2012
.
Techniques for telemetry transmitter attachment and evaluation of transmitter effects on fish performance
. Pages
45
87
in
Adams
NS
,
Beeman
JW
,
Eiler
JH
,
editors.
Telemetry techniques: a user guide for fisheries research
.
Bethesda, Maryland
:
American Fisheries Society
.
Lum
J
,
Taylor
S
.
2006
Dolly Varden and cutthroat trout migrations at Auke Creek in 2003, and abundance of cutthroat trout in Auke Lake, southeast Alaska
.
Anchorage: Alaska Department of Fish and Game, Divisions of Sport Fish Research and Technical Services (see Supplemental Material, Reference S11, 10.3996/112014-JFWM-083.S17); also available: http://www.sf.adfg.state.ak.us/FedAidPDFs/fds06-09.pdf (January 2012)
.
Margolis
L
,
Groot
C
,
Clark
WC
,
editors
.
1995
.
Physiological ecology of Pacific salmon
.
Vancouver, Canada
:
UBC Press
.
Maslin
PE
,
McKinney
WR
,
Moore
TL
.
1998
.
Intermittent streams as rearing habitat for Sacramento River Chinook salmon
.
Unpublished report prepared for the U.S. Fish and Wildlife Service under the authority of the Federal Grant and Cooperative Agreement Act of 1977 and the Central Valley Improvement Act (see Supplemental Material, Reference S12, 10.3996/112014-JFWM-083.S18); also available: http://www.calwater.ca.gov/Admin_Record/D-022206.pdf (November 2013)
.
McDonough
OT
,
Hosen
JD
,
Palmer
MA
.
2011
.
Temporary streams: the hydrology, geography, and ecology of non-perennially flowing waters
. Pages
259
289
in
Elliot
HS
,
Martin
LE
,
editors.
River ecosystems: dynamics, management, and conservation
.
Hauppauge, NY
:
Nova Science Publishers
.
Morbey
Y
.
2000
.
Protandry in Pacific salmon
.
Canadian Journal of Fisheries and Aquatic Sciences
57
:
1252
1257
.
Mosher
KH
.
1968
.
Photographic atlas of sockeye salmon scales
.
Fishery Bulletin
67
:
243
280
.
Nelson
B
.
1993
.
Movements of radio-tagged sockeye salmon in Auke Lake, southeast Alaska
.
National Marine Fisheries Service Auke Bay Laboratory. Unpublished Draft Technical Memorandum
(see Supplemental Material, Reference S13, 10.3996/112014-JFWM-083.S19)
.
Newell
J
.
2005
.
Migration and movement patterns of adult sockeye salmon (Oncorhynchus nerka) in Lake Washington
.
Master’s thesis. Seattle: University of Washington (see Supplemental Material, Reference S14, 10.3996/112014-JFWM-083.S20); also available: http://weber.s.uw.edu/research/publications/ms_phd/Newell_J_MS_Sp05.pdf (October 2013)
.
Newell
JC
,
Quinn
TP
.
2005
.
Behavioral thermoregulation by maturing adult sockeye salmon (Oncorhynchus nerka) in a stratified lake prior to spawning
.
Canadian Journal of Zoology
83
:
1232
1239
.
Olsen
JC
.
1968
.
Physical environment and egg development in a mainland beach area and an island beach area of lliamna Lake
. Pages
169
197
in
Burgner
RL
,
editor.
Further studies of Alaska sockeye salmon
.
Seattle
:
Univ. Wash. Publ. Fish. New Ser. 3.
.
Quinn
TP
.
2005
.
The behavior and ecology of Pacific salmon and trout
.
Seattle
:
University of Washington Press
.
[RDCT] R Development Core Team
.
2013
.
R: A language and environment for statistical computing
.
Vienna
:
R Foundation for Statistical Computing
.
Roscoe
DW
,
Hinch
SG
,
Cooke
SJ
,
Patterson
DA
.
2010
.
Behaviour and thermal experience of adult sockeye salmon migrating through stratified lakes near spawning grounds: the roles of reproductive and energetic states
.
Ecology of Freshwater Fish
19
:
51
62
.
Thornton
TF
.
2012
.
Watersheds and marinescapes: understanding and maintaining cultural diversity among southeast Alaska Natives
. Pages
123
136
in
Johnston
BR
,
Hiwasaki
L
,
Klaver
IJ
,
Ramos Castillo
A
,
Strang
V
,
editors.
Water, cultural diversity, and global environmental change: emerging trends, sustainable futures?
Jakarta, Indonesia
:
United Nations Educational, Scientific and Cultural Organization and Dordrecht, The Netherlands: Springer
.
Varnavskaya
NV
,
Wood
CC
,
Everett
RJ
,
Wilmot
RL
,
Varnavsky
VS
,
Midanaya
VV
,
Quinn
TP
.
1994
.
Genetic differentiation of subpopulations of sockeye salmon (Oncorhynchus nerka) within lakes of Alaska, British Columbia, and Kamchatka, Russia
.
Canadian Journal of Fisheries and Aquatic Sciences
51
(
S1
):
147
157
.
Wentworth
CK
.
1922
.
A scale of grade and class terms for clastic sediments
.
The Journal of Geology
30
:
377
392
.
Wigington
PJ
Jr
,
Ebersole
JL
,
Colvin
ME
,
Leibowitz
SG
,
Miller
B
,
Hansen
B
,
Lavigne
HR
,
White
D
,
Baker
JP
,
Church
MR
,
Brooks
JR
,
Cairns
MA
,
Compton
JE
.
2006
.
Coho salmon dependence on intermittent streams
.
Frontiers in Ecology and the Environment
4
:
513
518
.
Young
DB
,
Woody
CA
.
2007a
.
Spawning distribution of sockeye salmon in a glacially influenced watershed: the importance of glacial habitats
.
Transactions of the American Fisheries Society
136
:
452
459
.
Young
DB
,
Woody
CA
.
2007b
.
Dynamic in-lake spawning migrations by female sockeye salmon
.
Ecology of Freshwater Fish
16
:
155
164
.

Author notes

Citation: Ray JDM, Sethi SA, Eiler JH, Joyce JE. 2015. Pre-spawning movements and spawning distribution of sockeye salmon in an urbanizing Alaskan lake. Journal of Fish and Wildlife Management 6(2):472–485; e1944–687X. doi: http://dx.doi.org/10.3996/112014-JFWM-083

The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.

Supplemental Material