Abstract

In Lake Washington, juvenile sockeye salmon (Oncorhynchus nerka) strongly prefer Daphnia over other prey, switching uniformly to Daphnia when the threshold abundance of 0.4 Daphnia L−1 is achieved. Using long-term Lake Washington data (1978–2001) and fry trap data (1992–2001) from a major tributary, we examined the following: (i) factors that predict Daphnia pulicaria and Daphnia thorata increase to this threshold “switching” abundance, (ii) trends in Daphnia dynamics that may affect sockeye foraging and (iii) temporal correspondence of Daphnia increase and fry arrival. The winter abundance of D. pulicaria, in combination with basic parameters of spring conditions, was an important predictor of the date of D. pulicaria spring increase, indicating greater reliance on pelagic population dynamics (versus diapause hatch) than D. thorata exhibited. In addition, D. pulicaria was a more consistent prey than D. thorata, the latter exhibiting larger population fluctuations. Thus, recently increasing D. thorata prominence could decrease diet consistency for sockeye fry. Additionally, the timing of sockeye arrival to Lake Washington and Daphnia’s increase to the switching threshold has become less concordant, so that fry in recent years have had to rely upon less profitable prey for longer periods. Long-term trends and species-specific differences in Daphnia phenology may affect fry through altering diet composition, with additional implications for other zooplankton withstanding greater predation pressure in Daphnia’s absence. Recent decades of warming in Lake Washington are consistent with the warming of lakes worldwide, and complex phenological responses such as those reported here may be common as the climate continues to change.

Received October 16, 2005; accepted in principle December 14, 2005; accepted for publication January 26, 2006; published online February 2, 2006
 Communicating editor: K.J. Flynn

INTRODUCTION

Food availability is vitally important to juvenile fish, to support growth to developmental stages that are competitively superior, less vulnerable to predators or both (Werner and Gilliam, 1984; Cushing, 1996). Growth and survival of young fish depend not only on absolute abundance of potential prey, but also on the right prey being in the right place at the right time (Cushing, 1990; Chick and Van Den Avyle, 1999). While early stages of fish can be opportunistic in prey selection, taxonomic composition (Chick and Van Den Avyle, 1999; Mazumder and Edmundson, 2002) and size distribution (Beaugrand et al., 2003) of the potential prey community critically influence fish productivity. Variability in spatial and temporal plankton abundance represents an evolutionary challenge to fish recruitment that depends on substantial overlap between young fishes and their primary prey. Such reliance on overlap implies that climatic effects on plankton phenology can negatively affect fish by destroying synchrony (i.e. predator–prey matching) (Beaugrand et al., 2003).

Spring peaks of zooplankton in lakes are typically strongly associated with those of their algal resources (Sommer et al., 1986) and are generated by massive emergences out of the diapause egg bank and from overwintering pelagic populations (Cáceres, 1998). The extent of zooplankton reliance on the diapause egg bank, versus instantaneous reproduction, for maintaining pelagic population abundance appears to be highly species-specific and can strongly influence dominance patterns in the zooplankton community (Cáceres, 1997, 1998). Diapause emergence is primarily triggered by environmental cues (temperature, light), such that large hatching events are apparent in the spring for some species, but hatching can also be quite variable throughout the growing season (Cáceres, 2001). Thus, it is not always clear whether a reliance on diapause for pelagic population maintenance will create a more or less predictable spring peak.

In Lake Washington, USA, juvenile sockeye salmon (Oncorhynchus nerka) strongly prefer Daphnia over other zooplankton taxa (Beauchamp et al., 2004; Scheuerell et al., 2005) and Daphnia typically dominates the zooplankton community biomass through spring and summer (Edmondson, 1994). Migration of wild and hatchery-reared fry into the lake precedes the increase in Daphnia populations, and fry feed on cyclopoid and calanoid copepods until Daphnia increase (Beauchamp et al., 2004). Even at the very low abundance of 0.4 Daphnia L−1, sockeye fry switch uniformly and almost entirely to feeding on Daphnia (Scheuerell et al., 2005). This strong preference for Daphnia corresponds to Daphnia’s generally high nutritional quality (Mills et al., 1989; Mayer and Wahl, 1997) and relative ease of capture (Drenner et al., 1978; Mills et al., 1984; Link, 1996) and is consistent with Mazumder and Edmundson’s (Mazumder and Edmundson, 2002) finding that juvenile sockeye salmon growth is more specifically dependent on Daphnia abundance than on coarser measures of lake productivity.

The primary Daphnia species in Lake Washington are Daphnia pulicaria and Daphnia thorata. Daphnia pulicaria is larger-bodied and historically has been numerically dominant. While both species become common in the spring, D. thorata displays much more variable dynamics and may not achieve peaks until later in the year (Fig. 1). Summer (May–August) populations of Daphnia in Lake Washington have declined somewhat in recent years (Winder and Schindler, 2004a), and these declines appear to be primarily associated with D. pulicaria (Fig. 1). Modest increases in D. thorata offset these declines such that average annual Daphnia abundance in the lake has remained stable (Scheuerell, 2004).

Fig. 1.

Abundance of Daphnia pulicaria (solid line) and Daphnia thorata (dashed line) across each year from 1978 through 2001. A horizontal line indicates the threshold for sockeye switching, 0.4 Daphnia L−1. Arrows indicate the median date for migration of juvenile sockeye salmon entering from the Cedar River in the time period for which data are available from 1992 to 2001, for both hatchery-reared (H) and wild (W) stocks.

Fig. 1.

Abundance of Daphnia pulicaria (solid line) and Daphnia thorata (dashed line) across each year from 1978 through 2001. A horizontal line indicates the threshold for sockeye switching, 0.4 Daphnia L−1. Arrows indicate the median date for migration of juvenile sockeye salmon entering from the Cedar River in the time period for which data are available from 1992 to 2001, for both hatchery-reared (H) and wild (W) stocks.

Factors that influence timing of Daphnia’s increase to the levels at which sockeye fry begin to focus on them are expected to indirectly affect sockeye growth in Lake Washington. Here we consider the Daphnia spring dynamics from the fish perspective, asking (i) what factors influence the date at which each Daphnia species increases to the level at which sockeye feed almost exclusively on them, (ii) how do intra-annual and long-term dynamics of the Daphnia species differ in ways that could influence sockeye foraging patterns and (iii) how closely does the timing of juvenile salmon arrival match Daphnia spring peaks in the lake?

METHODS

Prediction of Daphnia spring increase

Lake Washington, USA (47°63′N, 122°26′W) has been the subject of uninterrupted long-term study since 1962, with less continuous data available back to 1933 (Edmondson, 1994). Detailed methods for data collection are presented elsewhere (Edmondson and Lehman, 1981; Edmondson and Litt, 1982; Edmondson et al., 2003). While data are available from several sites on Lake Washington, those from 20 m to the surface at the central sampling station (Madison Park) provide the most continuous data set and only those data are analyzed here. Romare et al. (2005) noted spatial variation in plankton dynamics across Lake Washington and found the highest planktivore abundance at Madison Park, suggesting special relevance of Madison Park Daphnia data to sockeye foraging. We used Madison Park data from 1978 through 2001 for the analysis of Daphnia dynamics.

To determine the date at which abundances of D. pulicaria and D. thorata sustainably reached the threshold “switching” level for sockeye fry, we identified the sampling date at which abundance for each species equaled or exceeded 0.4 ind L−1 (sensuScheuerell et al., 2005) and remained above this level for at least 2 following weeks and then log-linearly interpolated abundance for prior dates. Date of threshold onset was the interpolated Julian date at which abundance was equal to or greater than 0.4 ind L−1. In no instance did the combined abundance of both Daphnia species reach this threshold before one of the species could exceed the threshold alone, so analyses for each species proceeded independently.

Factors considered in multiple regressions predicting date of threshold onset within each Daphnia species were the Spring Temperature, Spring Chlorophyll and Winter Abundance. Spring Temperature was the average water temperature in March of each year from 20 m to the surface, ranging from 6.7 to 9.4°C. Spring Chlorophyll was average March Chlorophyll a measured in surface samples for each year, ranging from 2.8 to 10.1 µg L−1. March was chosen as the representative spring value because this is the month in which diatom blooms typically begin (Arhonditsis et al., 2003), concurrent with the lake’s warming (Arhonditsis et al., 2004). We averaged January and February abundance of adults of each Daphnia species to obtain Winter Abundance. January and February values were used as representative winter months because the lake’s stratified season frequently extends into December (Winder and Schindler, 2004b).

Multiple regressions were executed in JMP 5.1 (SAS Institute). The performance of resulting models was compared using the Akaike Information Criterion (AIC), adjusted for relatively small sample size (AICc, Burnham and Anderson, 2002). We calculated the Durbin–Watson statistic for the best-fit models of each species, to evaluate the possibility that data points were temporally auto-correlated.

Timing of sockeye fry migration

In response to declines in returning sockeye salmon, a multi-agency effort was initiated in 1992 to evaluate survival rates for the early sockeye life stages in the Lake Washington system (Seiler et al., 2004). The Cedar River is the main tributary to Lake Washington and the main contributor of sockeye fry to Lake Washington from both wild spawners and hatchery production (Beauchamp et al., 2004). Migration timing from the Cedar River is related to in-stream temperature (Seiler et al., 2004), creating the possibility that fish and daphnid timing could respond to regional warming similarly. Cedar River sockeye fry were sampled using a downstream migrant scoop trap from January through June. Hatchery-reared fry were differentiated from wild fry by thermally-induced otolith marking (Volk et al., 1999) at the Landsburg Hatchery. Detailed collection and population estimation methods can be found elsewhere (Seiler et al., 2004, 2005). We examined median migration dates for wild and hatchery-reared fry in relation to Daphnia threshold onset dates in the 9 years for which the data sets overlap, using general linear models in JMP 5.1 (SAS Institute), and we used Durbin–Watson tests to evaluate the possibility for temporal autocorrelation to affect analyses.

RESULTS

Daphnia spring increase

All models for predicting the onset date of spring threshold densities of D. pulicaria were significantly better than the null model (Table I), but D. thorata’s date of threshold onset was not described well by any model (Table II). For D. pulicaria, the full model including Spring Temperature, Spring Chlorophyll and Winter Abundance explained 71% of the variance in the data, but with so few data points, the corrected AIC (AICc) suggested that this model was overparameterized. Models including Winter Abundance with Spring Chlorophyll or Spring Temperature thus were tied as the most parsimonious models for D. pulicaria (Table I), suggesting that either of these spring parameters in combination with Winter Abundance was sufficient to predict D. pulicaria’s increase date. In contrast, no model was significantly better for D. thorata than the null model (Table II). The Durbin–Watson test statistic did not indicate that temporal autocorrelation was a significant problem for the D. pulicaria (P = 0.244) or D. thorata (P = 0.424) data.

Table I:

Results of linear models predicting the date at which Daphnia pulicaria reached and sustained an abundance of 0.4 ind L−1 for at least 2 weeks

Daphnia pulicaria models SSE DFE MSE R2 P AIC AICc 
Null 8799 23 383 0.000  144 144 
Spring Chlorophyll + Winter Abundance 4094 20 205 0.535 0.001 131 133 
Spring Temperature + Winter Abundance 4209 20 210 0.522 0.002 132 134 
Winter Abundance 6056 22 275 0.312 0.005 138 139 
Spring Temperature + Spring Chlorophyll + Winter Abundance 2581 16 161 0.707 0.002 128 139 
Spring Chlorophyll 6531 22 297 0.258 0.011 139 140 
Spring Temperature + Spring Chlorophyll 5583 20 279 0.366 0.025 139 141 
Spring Temperature 7507 22 341 0.147 0.065 142 143 
Daphnia pulicaria models SSE DFE MSE R2 P AIC AICc 
Null 8799 23 383 0.000  144 144 
Spring Chlorophyll + Winter Abundance 4094 20 205 0.535 0.001 131 133 
Spring Temperature + Winter Abundance 4209 20 210 0.522 0.002 132 134 
Winter Abundance 6056 22 275 0.312 0.005 138 139 
Spring Temperature + Spring Chlorophyll + Winter Abundance 2581 16 161 0.707 0.002 128 139 
Spring Chlorophyll 6531 22 297 0.258 0.011 139 140 
Spring Temperature + Spring Chlorophyll 5583 20 279 0.366 0.025 139 141 
Spring Temperature 7507 22 341 0.147 0.065 142 143 

In the rows listed below the null model, models are ranked from best to worst according to corrected Akaike Information Criterion (AICc), the AIC value as corrected for small sample size.

Table II:

Results of linear models predicting the date at which Daphnia thorata reached and sustained an abundance of 0.4 ind L1 for at least 2 weeks

Daphnia thorata models SSE DFE MSE R2 P AIC AICc 
Null 28744 23 1250 0.000  172 172 
Spring Chlorophyll 26330 22 1197 0.084 0.490 172 173 
Winter Abundance 28114 22 1278 0.022 0.170 174 175 
Spring Temperature 28740 22 1306 0.000 0.952 174 175 
Spring Temperature + Winter Abundance 24207 20 1210 0.158 0.318 174 176 
Spring Chlorophyll + Winter Abundance 24790 20 1240 0.138 0.387 175 177 
Spring Temperature + Spring Chlorophyll 25578 20 1279 0.110 0.495 175 177 
Spring Temperature + Spring Chlorophyll + Winter Abundance 21883 16 1368 0.239 0.660 180 191 
Daphnia thorata models SSE DFE MSE R2 P AIC AICc 
Null 28744 23 1250 0.000  172 172 
Spring Chlorophyll 26330 22 1197 0.084 0.490 172 173 
Winter Abundance 28114 22 1278 0.022 0.170 174 175 
Spring Temperature 28740 22 1306 0.000 0.952 174 175 
Spring Temperature + Winter Abundance 24207 20 1210 0.158 0.318 174 176 
Spring Chlorophyll + Winter Abundance 24790 20 1240 0.138 0.387 175 177 
Spring Temperature + Spring Chlorophyll 25578 20 1279 0.110 0.495 175 177 
Spring Temperature + Spring Chlorophyll + Winter Abundance 21883 16 1368 0.239 0.660 180 191 

In the rows listed below the null model, models are ranked from best to worst according to corrected Akaike Information Criterion (AICc), the AIC value as corrected for small sample size.

Daphnia pulicaria clearly achieved higher abundances than D. thorata throughout the time series (Fig. 1), and these populations were also significantly more stable. The number of times that D. pulicaria crossed the 0.4 ind L−1 threshold each year (1.75 ± 0.85 annual threshold peaking events) was significantly less (P = 0.02) than that for D. thorata (2.54 ± 1.31 annual threshold peaking events). Examination of the differences between onset dates of the two species for each year shows that some phenological changes have occurred (Fig. 2). Prior to the 1990s, D. pulicaria reached the threshold density an average of 22 days before D. thorata. With the exception of notable outliers in 1993 and 1998, D. thorata now reaches the threshold 4 days before D. pulicaria on average.

Fig. 2.

Number of days between Daphnia pulicaria and Daphnia thorata reaching threshold abundances of 0.4 ind L−1. Date of threshold achievement (onset) for each species was determined to be the Julian date at which the threshold was reached and subsequently sustained for at least 2 weeks. The horizontal line is a zero line, where both species reach threshold abundance on the same day. Points below the zero line indicate years when D. pulicaria reached threshold first, and above the line are years when D. thorata first achieved and sustained threshold abundance.

Fig. 2.

Number of days between Daphnia pulicaria and Daphnia thorata reaching threshold abundances of 0.4 ind L−1. Date of threshold achievement (onset) for each species was determined to be the Julian date at which the threshold was reached and subsequently sustained for at least 2 weeks. The horizontal line is a zero line, where both species reach threshold abundance on the same day. Points below the zero line indicate years when D. pulicaria reached threshold first, and above the line are years when D. thorata first achieved and sustained threshold abundance.

Concordance of sockeye fry migration and Daphnia onset

Hatchery-reared fry always reached the lake before wild fry (Fig. 1). In the nine years for which both zooplankton and Cedar River fry data are available, sockeye fry preceded the increase of any Daphnia species at increasingly longer time intervals (Table III, Fig. 3). The period that sockeye juveniles wait for Daphnia (regardless of species identity, i.e. “total Daphnia” in Table III) to increase to the “switching” threshold has become significantly longer for wild salmon, and a similar increase in the waiting period is suggested by Fig. 3 but was not quite significant (P = 0.06) for hatchery-reared fish. This deterioration of temporal overlap is primarily attributable to significantly increasing differences between D. pulicaria peak timing and sockeye migration (Table III), as D. thorata’s temporal overlap with sockeye has not significantly changed over this time period (Table III).

Fig. 3.

Length of time period (d) that hatchery-reared and wild salmon from the Cedar River were present in Lake Washington before Daphnia of any species achieved and sustained a threshold switching density of 0.4 ind L−1 (Daphnia onset). Corresponding statistics are presented in Table III as “Total Daphnia”.

Fig. 3.

Length of time period (d) that hatchery-reared and wild salmon from the Cedar River were present in Lake Washington before Daphnia of any species achieved and sustained a threshold switching density of 0.4 ind L−1 (Daphnia onset). Corresponding statistics are presented in Table III as “Total Daphnia”.

Table III:

Results of regression analyses examining the length of time every year that sockeye juveniles were present in Lake Washington before Daphnia achieved the threshold abundance for sockeye switching, 0.4 ind L−1 in the lake

 Hatchery-reared salmon fry
 
   Wild salmon fry
 
   
 F-Ratio DFE R2 P F-Ratio DFE R2 P 
Daphnia pulicaria 5.0 0.39 0.05 7.4 0.48 0.03 
Daphnia thorata 0.99 0.11 0.35 0.6 0.07 0.44 
Total Daphnia 4.8 0.38 0.06 7.43 0.48 0.03 
 Hatchery-reared salmon fry
 
   Wild salmon fry
 
   
 F-Ratio DFE R2 P F-Ratio DFE R2 P 
Daphnia pulicaria 5.0 0.39 0.05 7.4 0.48 0.03 
Daphnia thorata 0.99 0.11 0.35 0.6 0.07 0.44 
Total Daphnia 4.8 0.38 0.06 7.43 0.48 0.03 

Each species was examined alone and then for the earliest arrival of either species (“Total Daphnia”). Trends are shown in Fig. 3.

DISCUSSION

The increase of D. pulicaria to levels that attract sockeye predation was readily predicted by spring conditions and its overwintering population levels, but D. thorata displayed dynamics that could not be predicted by our models. In 3 years of studying Lake Washington Daphnia across all of the primary sampling stations, Romare et al. (2005) also found that overwintering Daphnia populations were significant predictors of Daphnia spring increases. In the present study, the interspecific difference in predictability of spring increases suggests that of the two species, D. pulicaria abundance in Lake Washington relies more on overwintering pelagic populations, versus egg bank dynamics, than does D. thorata. Similarly, Cáceres (Cáceres, 1997) described the persistence of D. pulicaria in Oneida Lake as being more heavily influenced by pelagic dynamics than were those of its primary competitor Daphnia galeata.

From the fish perspective, greater reliance on pelagic dynamics makes D. pulicaria a demonstrably more predictable resource than D. thorata. Consistent with our inability to predict the spring increase of D. thorata based on overwintering individuals and environmental conditions, the annual dynamics of D. thorata across the time series were significantly more variable. While D. thorata frequently rose above the 0.4 ind L−1 threshold, levels did not stay above the threshold as consistently as did D. pulicaria’s.

During time periods when Daphnia abundance drops below the threshold, sockeye switch back to feeding on copepods (Scheuerell et al., 2005). In general, copepods are less profitable prey for freshwater fish than are Daphnia (e.g. Mayer and Wahl, 1997), based at least partly on greater difficulty in capturing copepods (Drenner et al., 1978; Mills et al., 1984; Link, 1996). Not surprisingly, Daphnia-dominated food webs are frequently reported to have much stronger bottom-up effects on fish than those in which Daphnia are more rare (Mazumder and Edmundson, 2002). In Lake Washington thus far, at the times when D. thorata dropped below the threshold, D. pulicaria was generally abundant enough that sockeye fry would not have switched back to feeding on copepods. However, recent changes in the Daphnia species’ abundance and timing caution us that fry would encounter more variable Daphnia abundance, should dominance shift primarily to D. thorata.

From 1992 through 2001, sockeye fry have experienced increasingly longer periods of development in the lake before Daphnia could become the primary food source. This increasing mismatch is attributable to a trend toward later achievement of threshold levels for both Daphnia species, but especially for D. pulicaria. In this time period, March chlorophyll declined rather strongly (Scheuerell, 2004) for reasons that are unclear, but may have led to a delayed spring growth for D. pulicaria through effects on winter population members.

Of the three variables that explain 71% of variability in D. pulicaria’s onset timing, March chlorophyll is the only one that shows a negative trend that could affect D. pulicaria. Onset of stratification is an average of 16 days earlier in the spring (Winder and Schindler, 2004b), and the phytoplankton community is exhibiting complex responses that are currently under investigation (M. González-Sagrario, University of Washington, personal communication). While the declining trends in March chlorophyll and delayed Daphnia growth were relatively strong during the 1990s, the relationships are not as evident when the entire time series is inspected (S. E. Hampton, University of Idaho, unpublished results), suggesting either that oscillations occur over longer time scales or that very recent changes in timing are still swamped by historical variance when all data are considered simultaneously.

Winder and Schindler (Winder and Schindler, 2004a) reported that an increasing mismatch between earlier diatom peaks and comparatively stationary Daphnia peaks has led to declines of Daphnia’s summer (May–August) abundance. That work would appear to conflict with the present results, in that Winder and Schindler (Winder and Schindler, 2004a) reported that Daphnia peak timing has not changed; this seeming conflict is easily resolved by considering that those authors worked with total Daphnia abundance, while we attempted to discern differences in phenology at the species level. It is apparent that D. thorata’s increases in spring months offset D. pulicaria’s decreases (Figs 1 and 2) such that timing may appear stationary for total Daphnia, especially as relative to the diatom blooms in Lake Washington. Further, the present analyses suggest that the modern declines of Daphnia in May–August reported by Winder and Schindler (Winder and Schindler, 2004a) are more strongly affected by changes in winter survivorship than previously recognized; as such, the timing of total Daphnia increase is potentially more flexible than it would be if it were primarily the product of a spring diapause emergence with fixed timing.

CONCLUSIONS

Daphnia pulicaria has provided a predictable and consistent food base for the growth of sockeye fry in Lake Washington since its establishment in 1976 (Scheuerell et al., 2005). As a daphnid, it is characteristically nutritious and easy to capture and is also unique among its Lake Washington congeners in that the timing of its spring increase is easily predicted by overwintering population abundance and just a few basic parameters of the spring environment. Changes in abundance and timing of the two main Daphnia species in Lake Washington suggest growing prominence of D. thorata in Lake Washington, which would decrease the reliability of daphnid resources for salmon fry, should dominance ultimately switch to D. thorata, thus decreasing forage quality. Our study suggests that the potential for similar trends of mismatch to emerge in other Daphnia-dominated systems will depend on the degree to which Daphnia increases are related to pelagic (versus egg bank) dynamics. As shown here, such timing changes for Daphnia may be non-intuitive, depending upon phenological and abundance changes within the phytoplankton community that fuels daphnid growth in the spring. Recent reports of long-term warming in Lake Washington (Arhonditsis et al., 2004; Winder and Schindler, 2004b) are consistent with the warming of lakes worldwide (Magnuson et al., 2000), and complex phenological responses to temperature changes have the broad potential to disrupt communities.

ACKNOWLEDGEMENTS

The long-term Lake Washington data set is the result of the foresight and tenacity of the late W. T. Edmondson over his long career. D. E. Schindler provided many comments on the project and manuscript. J. M. Scheuerell’s work inspired this study, and her technical assistance and comments on all aspects have greatly improved it. We thank L. R. Nagy for advice regarding model selection and the numerous people that have contributed to the Lake Washington data set, particularly A. H. Litt and S. E. B. Abella. The Andrew W. Mellon Foundation currently supports Lake Washington data collection and archiving, and a National Science Foundation Post-doctoral Fellowship in Biological Informatics supported S.E.H.

REFERENCES

Arhonditsis
,
G.
, Brett, M. T., DeGasperi, C. L. et al. (
2004
) Effects of climatic variability on the thermal properties of Lake Washington.
Limnol. Oceanogr.,
 
49
,
256
–270.
Arhonditsis
,
G.
, Brett, M. T. and Frodge, J. (
2003
) Environmental control and limnological impacts of a large, recurrent spring bloom in Lake Washington.
Environ. Manage.,
 
31
,
603
–618.
Beauchamp
,
D. A.
, Sergeant, C. J., Mazur, M. M. et al. (
2004
) Spatial-temporal dynamics of early feeding demand and food supply for sockeye salmon fry in Lake Washington.
Trans. Am. Fish. Soc.,
 
133
,
1014
–1032.
Beaugrand
,
G.
, Brander, K. M., Lindley, J. A. et al. (
2003
) Plankton effect on cod recruitment in the North Sea.
Nature,
 
426
,
661
–664.
Burnham
,
K. P.
and Anderson, D. (
2002
)
Model Selection and Multi-Model Inference
 . Springer Verlag.
Cáceres
,
C. E.
(
1997
) Temporal variation, dormancy, and coexistence: a field test of the storage effect.
Proc. Nat. Acad. Sci. U S A.,
 
94
,
9171
–9175.
Cáceres
,
C. E.
(
1998
) Interspecific variation in the abundance, production, and emergence of Daphnia diapausing eggs.
Ecology,
 
79
,
1699
–1710.
Cáceres
,
C. E.
(
2001
) How well do laboratory experiments explain field patterns of zooplankton emergence?
Freshw. Biol.,
 
46
,
1179
–1189.
Chick
,
J. H.
and Van Den Avyle, M. J. (
1999
) Zooplankton variability and larval striped bass foraging: evaluating potential match/mismatch regulation.
Ecol. Appl.,
 
9
,
320
–334.
Cushing
,
D. H.
(
1990
) Plankton production and year-class strength in fish populations – an update of the match–mismatch hypothesis.
Adv. Mar. Biol.,
 
26
,
249
–293.
Cushing
,
D. H.
(
1996
)
Towards a Science of Recruitment in Fish Populations
 . Excellence in Ecology 7. Ecological Institute, Oldendorf.
Drenner
,
R. W.
, Strickler, J. R. and O’Brien, W. J. (
1978
) Capture probability: the role of zooplankter escape in the selective feeding of planktivorous fish.
J. Fish. Res. Board Can.,
 
35
,
1370
–1373.
Edmondson
,
W. T.
(
1994
) Sixty years of Lake Washington: a Curriculum Vitae.
Lake Reserv. Manage.,
 
10
,
75
–84.
Edmondson
,
W. T.
, Abella, S. E. B. and Lehman, J. T. (
2003
) Phytoplankton in Lake Washington: Long-term changes 1950–1999.
Arch. Hydrobiol. Suppl.,
 
139
,
1
–52.
Edmondson
,
W. T.
and Lehman, J. T. (
1981
) The effect of changes in the nutrient income on the condition of Lake Washington.
Limnol. Oceanogr.,
 
26
,
1
–29.
Edmondson
,
W. T.
and Litt, A. H. (1982) Daphnia in Lake Washington.
Limnol. Oceanogr
 .,
27
,
272
–293.
Edmondson
,
W. T.
and Litt, A. H. (
1987
) Conochilus in Lake Washington.
Hydrobiologia,
 
147
,
157
–162.
Link
,
J.
(
1996
) Capture probabilities of Lake Superior zooplankton by an obligate planktivorous fish – the lake herring.
Trans. Am. Fish. Soc.,
 
125
,
139
–142.
Magnuson
,
J. J.
, Robertson, D. M., Benson, B. J. et al. (
2000
) Historical trends in lake and river ice cover in the Northern Hemisphere.
Science,
 
289
,
1743
–1746.
Mayer
,
C. M.
and Wahl, D. H. (
1997
) The relationship between prey selectivity and growth and survival in a larval fish.
Can. J. Fish. Aquat. Sci.,
 
54
,
1504
–1512.
Mazumder
,
A.
and Edmundson, J. A. (
2002
) Impact of fertilization and stocking on trophic interactions and growth of juvenile sockeye salmon (Oncorhynchus nerka).
Can. J. Fish. Aquat. Sci.,
 
59
,
1361
–1373.
Mills
,
E. L.
, Confer, J. L. and Ready, R. C. (
1984
) Prey selection by young yellow perch: the influence of capture success, visual acuity, and prey choice.
Trans. Am. Fish. Soc.,
 
113
,
579
–587.
Mills
,
E. L.
, Pol, M. V., Sherman, R. E. et al. (
1989
) Interrelationships between prey body size and growth of age-0 yellow perch.
Trans. Am. Fish. Soc.,
 
118
,
1
–10.
Romare
,
P.
, Schindler, D. E., Scheuerell, M. D. et al. (
2005
) Variation in spatial and temporal gradients in zooplankton spring development: the effect of climatic factors.
Freshw. Biol.,
 
50
,
1007
–1021.
Scheuerell
,
J. M.
(
2004
)
Foraging Behavior of a Pelagic Predator and Long Term Shifts in Dominance Between Coexisting Daphnia Species
 . M.S. Thesis, University of Washington, Seattle.
Scheuerell
,
J. M.
, Schindler, D. E., Scheuerell, M. D. et al. (
2005
) Temporal dynamics in foraging behavior of a pelagic predator.
Can. J. Fish. Aquat. Sci.,
 
62
,
2494
–2501.
Seiler
,
D.
, Volkhardt, G. and Fleischer, L. (
2004
)
Evaluation of Downstream Migrant Salmon Production in 2001 from the Cedar River and Bear Creek
 . Washington Department of Fish and Wildlife. Olympia, WA.
Seiler
,
D.
, Volkhardt, G. and Fleischer, L. (
2005
)
Evaluation of Downstream Migrant Salmon Production in 2003 from the Cedar River and Bear Creek.
  Washington Department of Fish and Wildlife. Olympia, WA.
Sommer
,
U.
, Gliwicz, Z. M., Lampert, W. et al. (
1986
) The PEG-model of seasonal succession of planktonic events in fresh waters.
Arch. Hydrobiol.,
 
106
,
433
–471.
Volk
,
E. C.
, Schroder, S. L. and Grimm, J. J. (
1999
) Otolith thermal marking.
Fish Res.,
 
43
,
205
–219.
Werner
,
E. E.
and Gilliam, J. F. (
1984
) The ontogenetic niche and species interactions in size-structured populations.
Annu. Rev. Ecol. Syst.,
 
15
,
393
–425.
Winder
,
M.
and Schindler, D. E. (
2004
) Climate change uncouples trophic interactions in an aquatic ecosystem.
Ecology,
 
85
,
2100
–2106.
Winder
,
M.
and Schindler, D. E. (
2004
) Climatic effects on the phenology of lake processes.
Global Change Biol.,
 
10
,
1844
–1856.

Author notes

1University of Idaho, Department of Fish and Wildlife, Moscow, ID 83843, USA, 2University of Washington, School of Aquatic and Fishery Sciences, Seattle, WA 98105, USA, 3University of Lund, Department of Limnology, Lund, Sweden and 4Washington Department of Fisheries and Wildlife, 600 Capitol Way North, Olympia, WA 98501, USA