In several recent posts we have discussed the concept of density dependence and how it is used in fisheries management. Today we dive in deeper and talk about the stock-recruitment relationship, density dependence, and how the results of such models are applied to managing steelhead.
First, let’s define some terms. Stock refers to, in this context, a population of adult steelhead. We are generally most concerned with these fish once they return to freshwater since that is where we fish for them, where we enumerate (count or estimate) them, and where they spawn.
Recruitment refers to the number of progeny from a given year class, or cohort, which are recruited to the next generation of adults. Essentially, recruits are those fish which survive to adulthood and return in several years to spawn and repeat the cycle.
When these two parameters are plotted against one another they form what is known as a stock-recruit curve, the shape of which is a graphical representation of the density dependence concept. When we look at the lower part of the curve we see that as the number of spawners (stock) increases we get more recruits into the population. Makes sense, right? As we get further up the curve, however, more spawners results in a leveling off of the recruits — and in some cases the number of recruits can actually decrease at high numbers of spawners.
Fig. 1. The Beverton-Holt stock recruit function is represented by the red line, the gray area is a range of possibilities produced by the model and the data points are of Skagit River steelhead. Figure taken from the Skagit River Steelhead Fishery Resource Management Plan.
The graph is essentially showing the effect of density dependence on the population. Think about it this way: a stream is like a cafeteria with a limited amount of food. If there is a surplus of food you can keep adding individuals to eat there and everyone will be well fed. But as the number of cafeteria diners continues to grow, at some point the food supply will not support all those mouths. Consequently, some fish will starve or not grow as well, and eventually perish. In fact, at really high densities a bunch of fish competing for limited food and space can actually restrict growth of all individuals, resulting in lower growth rates for nearly all the population. This can have negative effects on recruitment.
Another important term in this analysis is the equilibrium point — the point at which recruits per spawner equals 1, or the population is stable. Beyond this point the productivity of the population stays flat or declines. The underlying assumption is that any number of fish which fall to the right of that equilibrium point will have little, if any, contribution to the population and in fact, could depress the abundance in future years. These are the fish typically defined as the “harvestable surplus” in fisheries management because removing them from the population would, in theory, not reduce abundance in future years. This concept of harvestable surplus based on a stock-recruit model is widely known as Maximum Sustained Yield, commonly shortened to MSY.
The approach has worked relatively well with sockeye salmon, which have a relatively simple set of life histories. However, a fundamental assumption with these types of models is that harvest will impact life histories equally. We can have confidence in this assumption when it is applied to sockeye, where most fish do the same thing, but not necessarily when it is applied to steelhead, which are much more diverse and iteroparous (characterized by multiple reproductive cycles over the course of its lifetime). With steelhead it is much easier to “mine” life histories disproportionately, thereby adversely affecting or even eliminating important life histories or components of the run unintentionally. Further, unless there is adequate monitoring of population structure this loss of life histories can easily go unnoticed.
Another issue is that managers assume the stock-recruit curve and its density dependent relationship accurately depict the capacity of a given watershed to support steelhead. As we reviewed in earlier posts (see them here), recent research suggests this is problematic for fishes because the occurrence of density dependence, and the equilibrium point in a graph, can occur even when populations are at very low abundance levels.
In other words, we still see density dependence in highly depleted populations where the habitat has not changed in quality. Consequently, while stock-recruit curves help inform us about how productivity changes with abundance of adults, they do not tell us about how much of the watershed is being used. Nor do they inform us about the potential capacity of the watershed, particularly with species like steelhead that display unique life histories that utilize different parts of the watershed at different times. Accordingly, we should be cautious about how we interpret these models when implementing fishery management plans for steelhead populations that have already been depleted.
We can continue to fish for steelhead while wild fish populations are recovering if we are careful in their management. We must allow enough fish to colonize underutilized habitat and rebuild depleted life histories. Which is why, in places like the Skagit River where wild steelhead runs have recovered enough to justify a fishery, we support a conservative approach which will ensure they continue to thrive.