The lazy fish: Chingik Dolly Varden retire early

See the paper here, or email me for a reprint


Anadromous migrations are one of the most recognizable characteristics of salmon. Young salmon head to sea for rearing where productivity and growth potential is higher than in freshwater environments. Semelparous salmon (e.g. Chinook, coho, sockeye, chum, pink) return to freshwater after a year for spawning and death after a year or more at sea. However, the pattern of migration and maturity is complicated in the iteroparous, facultitvely anadromous species (bull trout, cutthroat trout, Dolly Varden, brown trout, Arctic char, steelhead, etc.) that may vary widely in age at ocean migration and duration of time at sea, even within a population. In some cases individuals may remain in fresh water their entire lives, forgoing any ocean migrations.

Dolly Varden, Salvelinus malma, are a char species (closely related to Arctic char, Lake trout, brook trout and others) that is often facultatively anadromous where ocean access is available. Like many char species, they seem to have short ocean migrations only lasting the few summer months. In Chignik Lakes, Alaska, they are found in nearly every habitat, but migratory individuals primarily enter saltwater shortly after ice out (~ May), remaining there until August or September when the move back upstream for spawning or overwintering. Given this, in midsummer we would expect to find larger, older Dolly Varden in marine waters, and the smallest, youngest ones in fresh water. However, when we set about sampling Dolly Varden throughout the watershed, the very largest,oldest individuals were found in the Alec River, a headwater river about 25 km from marine waters. This begged the question: Why weren’t these fish at sea with the others?

The Alec River, Chignik Lakes, Alaska

The Alec River, Chignik Lakes, Alaska

We came up with several hypotheses about why we might find this pattern. First, Dolly Varden in Chignik may exhibit two distinctly different life histories, some individuals are migratory and others are resident. Like steelhead and rainbow trout, it seemed plausible that the Alec River Dollies were simply resident fish that never went to sea.
Second, it is possible that the large fish in the Alec River were either males or females remaining in fresh water while the other sex was migratory. In some Atlantic salmon populations nearly all of the migratory fish are female, while males remain in fresh water. This is because in females egg number is tied to body size, and they may have much more incentive to go to sea and get big.

Third, it could be that all fish are migratory, but not in every year. Therefore, the individuals we observe in the Alec River are simply those that skipped migrating for a season before resuming.

Fourth, fish are migratory at young ages, but cease migrating as they age.

To determine which hypothesis was supported, we needed a way to evaluate the migratory history of many individuals. For this we employed otolith microchemistry. Otoliths (ear stones) form a permanent record of the water chemistry that fish encounter. Marine chemistry differs from that of fresh water, and is detectable in the otoliths. We scanned the otoliths of fish from throughout the system to compare the migratory histories of individuals of many ages. What we found was that most fish in the system make at least one migration to sea. However, many of them make one or two consecutive migrations then effectively “retire,” remaining in fresh water for the duration of their lives. Fish begin retiring at about age 4, and by age 6 most fish are retired. Additionally, as we might expect, fish that fish that go to sea are larger for their age than those that remain in fresh water. But, that difference disappears by age 4, the age at which fish start retiring. What this means is that the benefits of marine migration are relatively weak in this system. We don’t know why this is exactly, but Chingik is a very productive system, and salmon subsidies are large for Dollies remaining in freshwater. Our previous work demonstrated that larger fish can survive and reproduce on a seasonal diet of salmon eggs alone.

An old, retired, Alec River Dolly Varden

An old, retired, Alec River Dolly Varden

So, it appears that Dolly Varden go to sea when they are too young (small) to survive on salmon subsidies, entering risky marine waters for a rich diet of sand lance, amphipods and crab larvae. Then, once they are large enough to survive on eggs, they retire in the relative safety of the Alec River and wait for the salmon eggs to come to them. It may be that fish from other habitats or tributaries with fewer salmon continue migrating, but we don’t have those data yet. To our knowledge, this is one of the first examples of a fish initiating migratory behavior, then later abandoning it for a sedentary life.

Resident Chinook in Alaska?


Lake Creek Chinook captured angling (Scott Slater)

Lake Creek Chinook captured angling (Scott Slater)

See the paper here, or e-mail me for a PDF.

A few years ago, we learned about a fishing lodge on Lake Creek, a tributary of the Yentna River, which drains into Cook Inlet, Alaska. For years the guides and clients at the lodge had observed a curious phenomenon; small, male Chinook salmon feeding on salmon eggs weeks or months before larger Chinook arrived to spawn. Anglers are in general keen natural historians, and they recognized that the Chinook they were seeing may have been non-anadromous, which would be essentially undocumented in a natural system with ocean access.

There were a few reasons why the anglers thought these might be resident fish:

  • They caught them really early in the season. Weeks before any large anadromous fish were in the river, these fish could be found.
  • They were all males, and some were quite small (< 300 mm fork length or ~11.5″), which seemed smaller than the jacks they were expecting to see.
  • They were oddly colored; many were a deep olive green with few individuals looking like bright, recent arrivals from the ocean.
  • They were aggressive feeders. Anglers regularly caught these little Chinook while fishing for rainbows. Scott has even caught them on dead drifted nymphs and even mouse patterns!

Tom Quinn and I collaborated with Scott Slater, a guide at the lodge, to initiate a small research program to investigate a potentially unique Chinook life history. Scott began collecting some basic body size data, and archived otoliths and muscle tissue. The otoliths were used to determine if the fish had ever been to sea by analyzing the ratio of strontium to calcium at different points along a transect from the core (juvenile) to the edge (recent). Sr/Ca is generally high in the ocean, and low in most fresh waters (the Yentna is no exception). Similarly, we analyzed the stable isotopes of carbon and nitrogen in the issues, as these can also be indicative of life in marine or fresh water.

Along the way, Scott also kept track of what was in the gut of these fish. The fact that these fish were readily taking egg patterns while people were trying to angle for rainbow trout was one of the first indicators that they might be resident fish. Scott found lots of invertebrates in the guts which is curious for semelparous salmon returning from sea.


Inverts found in the gut of a small Lake Creek Chinook. (Scott Slater)


More invertebrates in another small Chinook gut. (Scott Slater)

So, what did we find?

  • Stable isotopes indicated a pretty strong marine signal. We compared these to large, clearly anadromous, fish, and found that the little guys were actually more enriched. This is probably not an indication that the small fish were feeding at a higher trophic level than the large fish, but rather, an indication of a more coastal distribution of a fish that is only at sea for a year or less. Still, it is possible that as egg eaters, they could have a marine stable isotope signature without going to sea.
  • The otolith chemistry showed clear extended residence in waters with elevated Sr/Ca. In this system the Sr/Ca is very low in fresh water, so they would have to go to sea to pick up a signature like this. But, the signature was pretty variable, and might indicate a marine residence that was coastal and possibly estuarine.

Jack Chinook often return earlier than older age classes, but Lake Creek appears to have a very robust population of them, and they come back very early. How important their feeding is to their ecology remains to be seen. In addition to the unusual consumption patterns, we would like to know where these fish go at sea.

A few more photos of these little kings:


Lake Creek jack Chinook (Scott Slater)


Lake Creek Jack Chinook (Scott Slater)


Scott Slater with a slightly larger Chinook in Lake Creek

Comparing fish growth among habitats

If you have journal access, see the paper here: Journal of Fish Biology

Much of our work in fish ecology and conservation revolves around understanding the tradeoffs fish face by occupying alternative habitats. For migratory fishes we would like to know how a given environment influences key attributes like growth and survival. Ideally, we could compare growth and survival among habitats. However, measuring these attributes is laborious, often requiring extensive tagging and recapture to identify individuals through time. For numerous, highly mobile species, recapture rates of tagged individuals are often dismal. Therefore, a single capture method to measure growth would allow huge advances in our understanding and comparison of habitats. Until recently, a single capture growth measure remained elusive.

Recent advances in blood plasma hormone assays allows us to measure the hormone directly responsible to stimulating cell division and somatic growth, insulin-like growth factor 1 (IGF1). IGF1 has a number of roles, and is also involved in maturation, masking its growth effects in maturing fish. However, in immature fish numerous laboratory studies have demonstrated the efficacy of plasma IGF1 concentrations as a proxy for recent growth (primarily in fish length).

Extracting blood from an anesthetized Dolly Varden in the Chignik Lakes watershed. (Photo: J. Griffiths)

Extracting blood from an anesthetized Dolly Varden in the Chignik Lakes watershed. (Photo: J. Griffiths)

In facultatively anadromous species (e.g. cutthroat trout, bull trout, Dolly Varden, brown trout, etc.), that is, those that may vary widely in age at ocean migration, or may exhibit partial migration, we often wish to understand the growth potential of going to sea compared to staying in freshwater. In this study, we compared plasma IGF1 concentrations of fish captured in three estuarine sites with fish found in river and lake habitats. We also used otolith (earbone) chemistry to identify differences in residence time at different locations.

We were surprised to find different growth rates among estuarine locations. Dolly Varden are highly mobile fish capable of moving among these locations in a matter of hours (most are just a few km apart), yet otolith chemistry indicates these fish remain near their site of capture for at least several weeks. Moreover, these same sites become de-watered at low tide, so fish must repeatedly move into each site at high tide. Less surprising was that growth in estuarine environments was higher than lake sites at nearly all month-site combinations. However, there was no site effect in the lake, with equally low IGF1 at all sites.

Finally, we found that fish found in the river between the lake and estuary were of recent freshwater origin, with low growth, early in the season; gradually shifting to fish of likely ocean origin, with higher growth, later in the season.

In summary, this is one of the first studies to use IGF1 to measure growth in a wild fish, and demonstrates both the fine-scale residence and growth differences in an estuarine fish.

fine-scale Dolly Varden population structure in Chignik Lakes, Alaska

The paper is open access from Ecology and Evolution, click here to download it

It’s hard to appreciate just how many small streams, many unnamed, exist along the Alaska Peninsula. Dolly Varden are known to be early colonizers of these streams following deglaciation. However, Dolly Varden life histories often seem paradoxically unsuited for very small coastal streams. Dolly Varden, like many other char, head to sea for the first time at a fairly large size compared to other salmon. This means that they might rear in freshwater for several years before achieving a large enough size to smolt. So, very small streams may become inhospitable in winter months as flows decrease or waters freeze. In streams that drain to lakes, young Dolly Varden fry can move into the lake, or through it to a larger stream. What do Dolly Varden do in the myriad small streams that drain to saltwater?

Chignik2010_1 029

Spit Creek, draining to Chignik Lagoon (some locals call it “Old Lady Creek”) where Dolly Varden fry were found


Spit Creek from the ground at low tide. Spring flows are around 40 cfs.

Chignik2010_1 034

Whalers Creek, draining into Chignik Lagoon near Hume Point. Dolly Varden fry were also found here.

We hypothesized that Dolly Varden in small streams draining to saltwater might be adapted to smolt, or tolerate seawater, at the end of their first year of life, as streams dry or freeze in winter months. Although we did not have the capacity to directly measure this, we electrofished several small streams and found Dolly Varden fry using these habitats.


Electrofishing Waterfall Creek (our name). There is only about 100 m from the high tide mark to a large waterfall (impassable), yet we found young Dolly Varden fry here.

Two lines of evidence support our hypothesis. First, streams draining directly to saltwater were genetically distinct from those draining to lake or large river habitats. The level of distinction was surprising given that some lagoon streams are only separated from freshwater streams by 5 km or so. Second, we didn’t find the larger, older Dolly Varden that one might expect if they were rearing in these streams for multiple years.

More work is needed to know what exactly is driving the patterns that we have observed, but there is a strong implication of alternate life histories in lagoon streams have led to the observed genetic isolation. These processes may be prevalent outside of Chignik, as many, many, small streams dot the coastline throughout Southwestern Alaska.

Variation in Dolly Varden migration timing on multiple scales

If you have journal access, see the paper here or e-mail me for a reprint.

As part of our research into the diversity of anadromous behavior in Dolly Varden, we initiated a study of Dolly Varden movements (mainly ascent into freshwater in summer and fall) on three different spatial/temporal scales. First, we used acoustic tags to identify when fish were moving from lagoon to fresh waters in the Chignik system (see photo below). We found a very bi-modal distribution, with some fish entering freshwater in mid-summer, shortly after tagging. A second group of fish entered as late as November.


A second aspect of the study involved comparing upstream migration timing in the Chignik system to that of other watersheds throughout Alaska. In many watersheds weirs (see the Chignik Weir, the largest in Alaska, below), acoustic counting stations, or fish wheels are used to enumerate upstream passage of salmon. Sometimes Dolly Varden are also counted (Sadly, Dolly Varden are not counted, or are counted but never entered into databases or published in many watersheds where monitoring occurs) and we can learn something about the median date of upstream migration in those systems.


It turns out that although there is a general latitudinal cline in migration timing, with fish in higher latitude streams returning later than those in low latitudes. This may be because at low latitudes fish enter coincident with salmon for egg/flesh feeding opportunities, while at high latitudes they remain in marine waters for feeding for as long as possible.

An additional analysis of the few streams that we have long-term data for indicate that nearshore ocean temperatures are correlated with variation in return timing. In systems where fish return before the peak temperature, warms years tend to drive fish upstream early, while in streams where fish return after peak temperatures, fish tend to return later in the warm years.

Together these analyses paint a clearer picture of the complex migratory patterns of Dolly Varden that Bob Armstrong labeled as the “manager’s nightmare” in a 1984 paper.

The amazing, changing gut of the Dolly Varden

The amazing, changing gut of the Dolly Varden

Check out press coverage of the Dolly Varden gut study at: Seattle Times, Anchorage Daily News, and Nature World News

If you have journal access, see the paper here. Otherwise, read on about the study:

The Alaska Salmon Program has been working in the Chignik Lakes watershed in Southwestern Alaska, for over 50 years. Each summer a few graduate students and technicians make Chignik Lake their home for three months while they conduct research related to the fish and environment.Dolly VardenIn 2008 I started a new project that I hoped would explain some of the curious migratory behaviors of Dolly Varden (Salvelinus malma) in the Chignik system and elsewhere. Dolly Varden in Chignik are anadromous, meaning that they are born in freshwater and eventually move to the ocean for feeding and growth. Dolly Varden are thought to move to the ocean in their third or fourth year of life, and only remain there for the summer months, returning to freshwater  to overwinter.

In the June of 2009, we were investigating one of the main sockeye salmon spawning tributaries in the Chignik watershed, the Alec River (pictured), when we noticed something odd. There were lots of relatively large Dolly Varden in the river, exactly when we expected them to be at sea. Why were these fish in the river when they could be feeding at sea?Alec River, Chignik Before salmon arrive, in August, there is not much to eat in the river, especially for these bigger fish. Our assumption was that we would only find young Dolly Varden (~6″) that were feeding on insects prior to their migration to the ocean. Oddly, there did not appear to be any small fish there, only larger (16″-20″), presumably older fish.

In addition to being large and abundant, Chignik Dolly Varden in June were also really skinny (pictured). They looked like they were wasting away.Chignik 228 A check of their stomach contents confirmed that they weren’t eating much. Fellow graduate student Jonny Armstrong has been thinking a lot about subsidies and movement of fish, so we teamed up to delve deeper into the Alec River Dolly Varden, and why they did not migrate when most other fish appeared to do so. We know that for many fish migrating to sea is a risky proposition, and fish only do it if the net benefits (e.g., size, fecundity, weighed against risk of mortality) of moving exceed those of staying. Maybe the Alec River fish had found a way to survive and reproduce year after year without going to sea? We did some bioenergetics calculations that just didn’t make sense. The only thing the fish appear to eat is salmon eggs, and they eat lots of them. During spawning Dolly Varden might have several hundred eggs in their gut at one time. The fish get obese during this time and gain a lot of weight. But, our calculations indicate that they would still run out of energy by the middle of the next summer; not surviving until the salmon return in August.

Plasticity in gut size allows many organisms to change their energetic costs, a fact that has been well documented for birds and snakes among other organisms. We went back to the Alec River to see if changing the size of the gut was allowing large bodied Dolly Varden to survive the lean time between egg pulses, while maximizing their egg intake in the short time while salmon spawn.

To see what we found, see these comparisons of two fish of the same length from June (Top, before eggs are available, 10 months after the last pulse of eggs), and late August (Bottom, after about 4 weeks of feeding on eggs:

spring dolly fall dolly

The fall fish is fat, but not just because it has a stomach full of eggs. The organs inside have changed too:

spring dolly guts fall dolly guts

A closeup of the gut (stomach, pyloric caeca, and intestine) show that the gut doesn’t just stretch out to accomodate eggs, the digestive machinery has gotten much larger (2-4x larger in mass).

 spring gutfall gut

So what  does this mean for the fish? Once Chignik Dolly Varden get to be near their full adult size, it appears that they can use the flexible gut strategy to remain in freshwater, acquiring enough resources to reproduce and live until the next bout of eggs. It may be that other species have similar strategies to deal with pulsed resources, particularly eggs, in the same way. However, many watersheds do not have nearly the egg subsidy they once had because of a reduction in population size, or eggs ending up in hatcheries rather than streams.