Learning to Swim: The Early Life of Steelhead and its Implications for Management Part One

By Nick Chambers

Steelhead possess a personality that any high school punk rocker kid would strive for.  Make a rule and they will break it, because let’s face it, they do whatever they want. Any effort to categorize them will only capture the average at best because they are all are really just doing their own thing. They come in early one year and on time the next, but always on their own schedule. They don’t listen to your prayers to the fish gods because it doesn’t matter much to them, and the more you want to find one the more they avoid you. This is what makes them so interesting to me, as soon as you think you have them figured out, they throw you a curveball. While this collective personality makes them the ultimate Pacific Northwest gamefish it also means we must manage them different from salmon who possess a narrower range of personalities, or life histories as scientists would call it. The expression of so much variation in steelhead is what allows them to use the widest variety of habitats of any of our anadromous salmonids, and in turn be abundant and resilient. Within a diverse population at least a few individuals are going to be highly productive even when there are droughts or floods, heatwaves or cold snaps, or whatever else nature can throw at them.

But like most animals, steelhead are not necessarily born with a devil may care attitude. They are small and vulnerable when they first emerge from the gravel. While they grow rapidly over the first month or two, many, if not most, will die during this period. This early life stage, known as fry, has received relatively little attention among recent research in steelhead but improving our understanding of it may be vital to their recovery.

To understand why this may be the case let’s take a step back.

First, in western Washington State, steelhead are generally managed using an escapement goal. This is based on the assumption that a given number of fish can use all the available habitat. If we let significantly more adults than goal spawn it is assumed that the resulting higher density of juvenile steelhead would lead to more competition for food and space and in turn reduced growth or survival, and ultimately, there would not be a meaningful increase in the number of returning adults in the next generation. This makes sense, there is only so much food and space available in the environment, so what is the problem here? To help answer that we can look halfway around the world to Atlantic Salmon and Brown Trout.

Their Sex Pistols to our Misfits.

Why Atlantic salmon? In this case, they are a better analog to Steelhead than Pacific Salmon due to the behavior of their juveniles (Bley and Moring 1988).  And the attribute we are most interested in here is the ability to disperse from the redd early in life. Researchers have found that the ability of Atlantic Salmon and Brown Trout fry to disperse is quite limited, with the majority of them staying within a few hundred meters of the redd during the first months of life (Einum et al. 2011). This time of high mortality and limited dispersal has been termed the critical period of survival; once they make it through the first one to two months of life, they have a better chance of surviving to the smolt stage. Importantly, the mortality is not just high, it is proportional to density (Milner et al. 2003). This means when redds are clustered together fry survive relatively poorly due to the increased competition for limited resources during the critical period. This is important because it appears that it is only after the critical period and enough growth has occurred that juveniles are able to disperse and alleviate the effects of competition.

To test how this might influence population growth, researchers artificially constructed redds using the same number of eggs in two different patterns. The first had redds clustered in two patches 200 meters apart, and the other had the same number of eggs distributed in multiple redds across 200 meters of a stream. In the sections with two clusters of redds, the fish survived poorly and were less evenly distributed and smaller compared to fish in the sections with evenly distributed redds (Einum et al. 2008).  Even when redds were spread out over short distances, the effects of competition were reduced and led to greater survival and growth.

This indicates that the distribution of spawning adults has important implications for understanding habitat capacity. When dispersal ability is limited and mortality is high, the competition that increases freshwater mortality at high densities occurs within the small patches of habitat near redds, even if there is ample habitat for later life stages in other parts of the watershed. If the redd distribution is patchy throughout the watershed, then fry may not be able to reach all the available habitat until after the period of greatest mortality is over. This means we could be underestimating the true capacity of the habitat by not accounting for how much of the habitat fry can actually reach early in life (Finstad 2013).

Some might argue that this effect of distribution is already accounted for in the models we use today, after all they generally compare one generation of adults to the next, so all the effects of distribution at younger life stages would show up at the adult stage, right? Well, it depends.

It is clear that our steelhead populations in the lower 48 are at a small fraction of their historical abundance, a result of overharvest, hatchery effects and habitat loss. There are two important things to note about the general effects to a population when declines in abundance occur.

First, as abundance decreases, the distribution of spawning fish across the landscape compresses into high-density core areas with a smattering of redds unevenly spread across the remaining habitat. This is fundamental to the study of ecology, what scientists would call a positive abundance occupancy relationship (Gaston et al. 2000). put, as abundance increases, a population will increase in both distribution and density.

The second is that diversity will be reduced. While salmon may only have a few life histories of fish, each represented by many individuals, steelhead has many life histories, with relatively few making up most of the strategies. When you have a lot of diversity, there is less redundancy in any one life history. Thus, life histories are relatively easy to reduce or eliminate through harvest and once they are gone, they can be difficult to rebuild.

For example, steelhead generally spawn over 6 months or so and we have seen reductions in the breadth of spawn timing, including on the Skagit (Meyers et al 2015) and Olympic Peninsula where early entering and spawning fish have been nearly lost (McMillan et al 2022). Importantly, the loss of these early fish occurred prior to 1980 in most populations and no obvious link to habitat loss exists to explain the elimination of just one population segment. A much more plausible solution is the high harvest rates targeted at the early return period where hatchery and wild fish were both abundant.

This suggests that the steelhead populations that remained by 1980 or so were unlikely to be at equilibrium with their environment. Equilibrium is science speak to say that the habitat is fully occupied, and when it is not at equilibrium there will be empty space and unexcused absences.

Enough background let’s bring it all together. Current estimates of habitat capacity (and corresponding escapement goals) are based on statistical models designed to detect the effects of juvenile density on trends in abundance, meaning that at some point in life there has been enough competition to limit the productivity of a population and keep it from growing beyond what the habitat can support. When this signal is observed in these models, capacity is assumed to have been reached and provides the basis for where we establish an escapement goal. But what the research we covered above is telling us is that for species that undergo a period of high mortality and limited dispersal ability early in life, these effects of competition can take place at just a single cluster of redds and are not necessarily reflective of what is occurring throughout the watershed. If the population in question has suffered previous declines in abundance unrelated to habitat (overharvest, hatchery effects, etc.) then we may be underestimating the true capacity of the habitat. And that habitat capacity signal? It may simply be coming from high-density core areas while other parts of the watershed are underutilized.

Stay tuned for Part 2 where we look at the early life of steelhead and if the exhibit the same trends as their Atlantic relatives.