Searching for Clues in Salmon Scales

I came to GMRI as a postdoctoral researcher to study how marine ecosystem changes were related to Atlantic salmon population trends. That research linked the abundance and productivity of Atlantic salmon populations across North America to ocean warming and associated changes to the food web, particularly a reduction of nutritious zooplankton and capelin (Mills et al., 2013). My lab and our collaborators are building on these findings by focusing on the growth process in Atlantic salmon at different points in their life cycles. Salmon are anadromous fishes, meaning they spend the early part of their lives in rivers, move into the ocean for one to several years as they grow and mature, and then return to their natal rivers to spawn. The consistent decline in populations spread over different rivers indicates that something is happening during their time at sea that is making it harder for them to survive. By studying growth patterns preserved on salmon scales, we can discern changes in salmon growth over time, estimate when those changes occurred during the fish’s time at sea, and then relate these findings to changes in ecosystem conditions and marine survival rates.

For example, previous work suggests that Atlantic salmon prefer ocean temperatures between 4-8 degrees Celsius and grow more slowly at the lower, colder end of that range compared to the higher, warmer end of that range. So, when we see groups of more narrowly spaced circuli, those were probably formed during the colder winter months or during periods of poor prey availability. In contrast, groups of circuli that are spaced further apart likely align with warm summer months and favorable feeding conditions. These initial insights help us better understand how seasonal conditions might have varied for a given fish, and help us align factors such as temperature or zooplankton abundance to the salmon’s growth during its time at sea.

The more we can understand how changes in the marine ecosystem are affecting Atlantic salmon, the easier it will be to identify the mechanisms behind these trends, and develop knowledge that can inform U.S. and international conservation and management efforts for Atlantic salmon.

Lindsay Gray Carlson, Miguel Barajas, and Mike Tillotson are three research associates on my lab team working to piece together this puzzle.

Whether it’s Pacific Black Brant Geese flying over 6,000 miles between breeding and wintering areas, or salmon migrating from rivers to the open ocean and back, I’ve always been fascinated by animal migrations. In my current research, I’m looking into how changing environmental conditions in the Northwest Atlantic ecosystem affect the growth, migration, and survival of Atlantic salmon. Tracking and studying salmon in the ocean isn’t as simple as it might seem. However, because salmon spend a critical portion of their lives in the ocean, it’s important for us to tackle this research challenge head-on, especially if we want to develop more effective conservation and management strategies for such an iconic species.

Most of what we know about where salmon go when maturing into adults in the open ocean comes from tagging efforts. Scientists and managers began putting numbered, plastic ID tags on adult Atlantic salmon returning to rivers in the Gulf of Maine, and eventually on hatchery-reared smolts (young salmon yet to migrate out to sea) in the 1960s. When Atlantic salmon were commercially fished in Canada and the North Atlantic, we could get information about their growth, migration timing, and distribution at sea because commercial fishers would collect data about the fish they caught, take scale samples, and return the tags and information. This data helped us understand where salmon go in the ocean at different times of the year and at different phases in their life cycle.

Based on these historical tagging efforts, we learned that after leaving their natal rivers, most North American Atlantic salmon migrate to the Labrador Sea, where they then spend the summer foraging for food and maturing into adults. In the winter, they move to the southern portion of the Labrador Sea. The salmon that reach sexual maturity at this point are known as “one-sea-winter (1SW) maturing” individuals, and they are the first to begin their migration back to their natal rivers in the following spring to spawn.

Most salmon will not mature during the first winter, and these individuals will spend at least one more year at sea before spawning. These salmon are known as “multi-sea-winter (MSW) maturing,” and these fish migrate to the west coast of Greenland in their second summer at sea. Researchers still don’t know exactly how long Atlantic salmon spend off of the Greenland coast, but due to the abundance of prey such as capelin, shrimp, and squid in this area, the longer migration is worth the trek.

Studies show that other long-distance migrants, such as seabirds, are disproportionately impacted by changing environmental conditions, and our lab is interested in how changing ocean and marine ecosystem conditions may be affecting the growth and survival of Atlantic salmon that migrate long distances at sea. However, because our understanding of marine migration timing and their relationship with scale patterns is limited, we can only coarsely link scale growth patterns to local environmental conditions and survival rates.

I’m working to improve our understanding of this relationship by figuring out when a distinct set of circuli — called the annulus — is completed. The annulus is one of a few highly distinguishable growth markers on a scale that could potentially be tied to a specific time in the life cycle of an Atlantic salmon, and to a specific place. However, our limited understanding of when and why the annulus forms makes it difficult to link growth trends with ecosystem conditions experienced by an individual salmon. I created an equation to estimate the date of annulus completion using data from the tagged salmon. This method works for tagged salmon because we know exactly how many days they were at sea, which allows us to calculate a seasonally variable growth rate. Based on these findings, we think that most individuals complete their annulus between mid-February and late March. Understanding more about the timing of salmon migration will help us understand the relationship between salmon growth and ecosystem conditions, especially as ecosystem conditions continue to change.

I studied anadromous fish like river herring and their ecological role linking rivers to oceans at the University of Southern Maine with Karen Wilson and Theo Willis. That work sparked my interest in researching animals that play a part in that interface between rivers and oceans, and working on Atlantic salmon at the Gulf of Maine Research Institute with Kathy Mills has been an exciting next step for me.

Both North American and European Atlantic salmon populations are dwindling, but for both regions, the magnitude of decline has been strongest in populations at the southern extent of the range and among population cohorts that remain at sea the longest before returning to the rivers to spawn. Many of these cohorts are from rivers in Maine, and no river has a larger population of Atlantic salmon in the United States than the Penobscot River.

Every year thousands of salmon smolts leave the Penobscot and less than a fraction of a percent successfully return. What happens to them? There are so many questions! Are the smolts eaten up by predators on their way out to sea? Which predators? Is there not enough food or suitable habitat? Is it too warm?

My research starts after all the field work and lab work has been completed and converted into data. More specifically, I look at the data that comes from scales of salmon that have successfully migrated back to the Penobscot River. These salmon that survive the return migrations are important because most salmon never return once they leave the river, perishing somewhere along the way instead. I work on writing code and algorithms that can detect and characterize changes in growth patterns between salmon cohorts, be it over time or between salmon that mature after one or two years at sea.

Among the salmon that manage to survive the marine migration back to the Penobscot, our analysis of scale growth patterns found evidence of reduced growth during the late marine phase — toward the end of their time at sea. In this final phase of growth, salmon no longer allocate energy toward getting bigger, but rather toward building up the reserves needed to make the long migration home. Penobscot River Atlantic salmon migrate farther south than any other North American Atlantic salmon population, and as such, spend more energy during their homeward migration compared to other salmon with shorter migrations. To add to that difficulty, they may have to begin their migration sooner than in the past, in order to reach their natal rivers while temperatures are optimal for spawning. That earlier departure to migrate home means they have less time for growth compared to their northern counterparts. In recent years, it appears that Penobscot River salmon are losing more of this last growth opportunity. During the same period these salmon have been returning earlier, they have also had lower marine survival rates. This raises questions about whether environmental conditions are spurring this earlier migration and what it may mean for the population long-term as the Gulf of Maine continues to warm.

This uncertainty is part of the reason why it’s so important to explore these questions about Atlantic salmon. If we want to protect them, we need to find some answers about what’s happening to them, and soon.

Like many fish biologists, my interest in studying fish grew out of time spent catching them. Growing up in Seattle, I would often fish for Pacific salmon. I had my first introduction to Atlantic salmon in Maine while attending Bowdoin College. The Androscoggin River runs through Brunswick, where the school is located, and I was interested in what fishing opportunities might be available in that river. Unfortunately, like most New England rivers, the legacy of dams and pollution left by industrial mills have created unfavorable environments for salmon. During my time at Bowdoin, fewer than 50 salmon returned to the Androscoggin each year, and in the most recent years, the counts have been in the single digits. It’s this troubling notion of empty rivers that has motivated my interest in understanding why there are so few salmon left in Maine and the Atlantic more generally.

I’m most fascinated by the way the environment and humans shape the diversity of salmon life history. With Maine’s populations dwindling and Atlantic salmon from most other regions faring only slightly better, the idea of commercial and recreational fishermen catching millions of salmon each year seems like a fantasy. However, salmon are not inherently fragile or vulnerable, and with suitable habitat and favorable environmental conditions, their populations can be remarkably productive. What has changed in the ocean that might be limiting this productivity, and how can growth patterns help us identify this change?

All of us working with Dr. Mills in the Integrated Systems Ecology lab are using scale archives to study growth patterns in Atlantic salmon, however, given the intricacies of our datasets and our analytical backgrounds, we’re all tackling this question in slightly different ways. The primary difference in my work is that instead of looking at Penobscot River data, my data comes from the fishery (once commercial, now subsistence only) that occurs each summer and fall off the coast of Greenland.

Salmon survival rates can change drastically depending on ecosystem variables, but it’s not clear which variables are most closely tied to their survival. I use modeling techniques to help narrow those variables down. I can use computing technology to piece together a lot of different information, such as growth pattern trends, survival rates, and age and stock composition through time. Then I can start to relate that information to potential explanatory variables (e.g., environmental conditions). Most people would assume that bigger fish have better chances of surviving in the ocean, which was also our lab’s hypothesis. To help figure this out, I decided to build on previous work that identified a big shift in North Atlantic ecosystem conditions around 1990 (when an influx of fresh water from the Arctic disrupted ocean stratification and zooplankton abundance) by comparing growth-survival relationships before and after that big shift. Surprisingly, I found that while bigger fish were related to increased survival rates before 1990, that no longer seems to be the case. This suggests that the very mechanisms regulating salmon survival may change over time, which has big implications for how we think about managing them. If a fish's size and survival aren't connected, we can't assume bigger fish means healthier fish populations. There are a lot of other potential explanatory variables out there, and narrowing those down is the basis of our future work.