John McMillan Podcast Clarification

Anchored Banner

*Note – after posting John’s podcast last week (Part 2), I received some great questions about our hatchery discussion (find the conversation here www.aprilvokey.com/podcasts/ ).  John responded to me with some point-form notes that he wished he’d mentioned during our initial conversation (which took close to two and a half hours… close to midnight).  I asked John to send me any clarification he wanted to share with all of you, and he kindly sent me the below message/text.  We hope it helps to answer some questions!  It’s such an in depth subject!!  ~April

From John:  “Ok, after listening to the podcast I realize I was all over the place in the discussion and that some statements were confusing. Most of that stemmed from me not defining that there are two types of hatcheries, and then often realizing – half-way through a topic – that I was using terminology that people might not understand. This is one reason that the hatchery issue, which is terribly complex, is difficult to discuss without being able to show data. So, I really appreciate the opportunity to provide clarification, because it is important for people to know what I was saying and to help others understand the main points of the discussion.  To that end, I outline what I think are the six biggest take-home messages on steelhead and hatcheries, and how we at TU seek to try to solve the debate. Again, all references are only to steelhead unless otherwise noted.

1. I thought I mentioned this in the podcast, but can’t recall, that there are two types of hatcheries. I made comments on both types, but our discussion ranged and it was not clear which we talked about at specific times. So, let’s clear that up. The two types of hatcheries are: segregated and integrated. Segregated use stock from one river and then spread it around to lots of rivers. Those fish are not meant to spawn in the wild and clearly have different genetics than the native stock. These are often referred to as “traditional” hatcheries because their main goal is to provide fish for harvest. Examples of these hatcheries include the Skamania stock of summer steelhead and the Chambers Creek stock of winter steelhead in Washington. Then there are integrated hatcheries, where they take the fish from the same basin, rear them in a hatchery, and then release them. These are often referred to as “wild broodstock” hatcheries.  Managers typically want these fish to spawn in the wild because they are from the same population.

2. Survival of the segregated hatchery fish is very poor, because not only does the environment of the hatchery generate maladapted behaviors (such as not being afraid of predators), but it has also selected for different genes because the hatchery only uses those fish as broodstock. So over time the hatchery population becomes very adapted to the hatchery environment. Hence, their poor survival in the wild, which is why we tend to see low levels of segregated hatchery genes in wild populations. These are the fish I talked about that often equate to a “zero” in survival. Now, don’t get me wrong, some segregated hatchery fish produce offspring that survive and we do see hatchery genes in wild populations, particularly where there are large releases of hatchery fish and few wild fish. And this has happened more frequently in research from California that I have seen. That said, I was only trying to answer the question of: Are there any wild fish left after planting hatchery fish for a 100 years?  I think the evidence indicates a strong “yes”, there are lots of wild fish left, largely because those traditional hatchery fish faired so poorly in the wild.

3.  Survival of the integrated fish is about 50% of wild fish, so better than the segregated stock, but not as good as wild fish. They thus produce more offspring that survive than segregated programs. Now, the integrated fish, essentially being wild fish, are also prone to effects of hatchery rearing and consequently, they also develop maladapted behaviors (such as not avoiding predators after their initial release) and are likely to have smaller brains, etc. But, research indicates that the phenotypic response is not really what we should be worried about the most. Instead, we should be concerned more about a genetic change. I discuss that below.

4. The question that I think was most confusing on my end was: How can hatchery fish be different from the wild fish if they come from the same stock?  Here I can clear up some of what I was trying to explain.

Let me step back, and reiterate that the hatchery rearing environment can change a number of aspects of the fish, including brain size, body shape, fin size, and otolith/lateral line makeup.  But, those are phenotypic changes that are induced by the environment of the hatchery and they are not necessarily passed along to offspring.  For instance, in my wolf-boy analogy a young man was raised with wolves and did not know how to communicate with humans. He did not speak and it would be likely that modern technology would find differences in his brain wiring compared to normal humans that have a long history of interacting with one another.  It would be the same as having a different region of the brain can take over other functions after injuries. Those types of changes are phenotypic because wolf-boy’s offspring would have all the capabilities of being a normal human.  In other words, it was his upbringing that changed him. His genetics did not change. And this gets into nurture v. nature. The bottom line is that things like small brain size are due to lack of stimuli in the hatchery, meaning that how the fish is nurtured instills changes, but the changes can be reversed in the next generation.

The phenotypic change is not the issue as much as the genetic change. Because if the changes in hatcheries were only phenotypic, then they would revert back to wild fish after spawning and rearing in nature. But, that is not what research is seeing for steelhead. For example, after first generation hatchery steelhead from wild stock spawned in the wild, their offspring also survived poorly (in Hood River studies). This suggests that the hatchery rearing affects genetics that are passed on generation-to-generation. Indeed, a follow up study found that those first generation hatchery steelhead differed genetically from wild steelhead in traits like immune system function and metabolism. They are all important, but I will focus on metabolism because it is something I am more familiar with.

Why is metabolism important and how might it be important to hatchery and wild steelhead survival?  Metabolism, which is also correlated with things like growth rate and behavior, appears to be the basis for many life history decisions in steelhead. For instance, one study recently showed that individuals that become steelhead are more likely to have faster, less efficient metabolisms than individuals that become rainbow trout. The same has been found in other salmon species too, with even more detail. In that research they found individuals with less efficient/faster metabolisms spend more energy on converting food to growth than those fish more efficient/slower metabolisms.  The latter can thus grow more quickly on limited food supplies than the former, while the former can grow more quickly when food supplies are high. All this means is that there is a genetic basis for metabolism and that metabolism has a strong influence on life histories. So we can essentially think of life histories as being a surrogate for metabolism.

So how could a hatchery select for a certain metabolism?  This is the crux of the issue. We don’t know for steelhead exactly, but, before we get there we need to understand that not all steelhead in a population have the same metabolism. Nor do they all have the same life histories. When thinking about selection in hatcheries we need to think in terms of distributions. For example, the distribution of individual metabolisms – or life histories –  can be thought of as a bell-curve. In other words, there is a distribution of individuals within a population ranging from – in most simple terms – slow to fast. We could also think of the distribution ranging from small steelhead, such as half-pounders, to large steelhead, such as 40lb 5-salt individuals.  Regardless, there is variation in steelhead and that diversity provides the basic material for selection in nature and hatcheries.

Metabolisms, and thus life histories, are strongly influenced by rearing environment, such as food supply and water temperature. Given that the hatchery is one relatively simple environment with lots of food and primo conditions for growth, it tends to select for a very simple set of life histories. The simple environment thus truncates the amount of diversity relative to the population that the fish were taken from.  Basically, it appears that hatcheries amplify certain life histories that do well in hatcheries, but less well in nature.  In this way there is less distribution/variation of life histories in hatchery fish than for wild fish, because there is less diversity in the hatchery.

Variation in life histories is critical to the survival and resilience of steelhead because nature is unpredictable and wild steelhead occupy a wide range of habitats. To persist in an unpredictable environment a species either becomes a specialist, like pink salmon where most fish do the same thing in the same place, or a generalist, like steelhead that do an array of things depending on the local environment. The life histories in steelhead allow them to spread risk of death across place and time and help dampen the negative effects of any single bad year or bad event in a given place. For example, studies show that wild steelhead populations have up to and over 30-38 life histories.  This is because smolts go to the ocean from age 1 to age 5 and spend weeks to 5 years in the ocean.  As mentioned, the hatchery produces a much simpler set of life histories because nearly all smolts migrate to the ocean at the same age and return to freshwater after spending 1-2 years in the ocean. We have essentially turned a biologically diverse fish into a pink salmon, so to speak.

So, yes they are the same species and from the same population is you use wild broodstock in a hatchery, but the hatchery simplifies their life histories and produces a reduced level of diversity. For this reason, the life histories and genetics of the hatchery population can differ from that of the wild population. Given the linkage between life histories and metabolism, it is my guess that we are probably simplifying the set of metabolisms too. For instance, faster metabolisms may be beneficial to fish in the hatchery where food is abundant because we want them to grow fast so that we can save on the cost of rearing. Fish with faster metabolisms are more aggressive in seeking out food and in competition with other individuals. While increased metabolic rates may help fish acquire more food in the crowded rearing conditions found in hatcheries, such a trait may be selected against in nature because it requires taking more risks to acquire food. More risks typically equals higher rates of predation. Most importantly, by having less diversity to draw upon because nearly all fish do the same thing, the hatchery fish become less able to survive in the wild and more prone to boom and bust cycles in survival.

5. Can we make hatchery fish more like, or even equal to, wild fish?  I hit on this a bit and there are two ways to look at this.  First, yes, we can make hatchery fish more like wild fish by rearing them in “enriched environments” where there is cobble and rock on the bottom of the tanks, there is flowing water, food is fed at a less predictable way and there are fewer fish per area.  We have research that indicates this makes the phenotype of the hatchery fish more closely resemble the phenotype of the wild fish, but we don’t know what it does to genetics, if anything.  The problem with this, as I explained in the interview, is that it is really expensive to rear fewer hatchery fish in enriched environments, so it is likely only a solution for non-production hatcheries. I don’t think it would be cost effective for production hatcheries. Further, it is probably physically and economically impossible to diversify the hatchery rearing environment to the point where it is producing multiple age classes of smolts that are consistent with those found in nature.  So, ultimately it is highly unlikely that we can make hatchery fish exactly like wild fish.

Second, it may be possible that nature can operate on wild broodstock hatchery fish to make them more like wild fish.  For instance, a big question is: If a hatchery fish spawns in the wild, will it’s offspring be the same as wild fish?  Well, we know that there is a genetic difference, so no, they are not the same if they are always reared in a hatchery and then released into nature. However, there is area for research here. For instance, we tend to run hatchery operations in perpetuity without stopping.  What we don’t know is if you run a hatchery program for a generation or two, then stop, what happens?  Some scientists are trying to look at this now, but the results are not in.  We know that not all hatchery fish survive worse than wild fish. Some individuals do better.  It is possible then that over time that those fewer hatchery fish that survive well could pass on their offspring and contribute positively to the population as a whole.  We don’t know at this time, but if possible, it would require changing the way we operate hatcheries.

6. Whew, hope some of you are still with me here and now understand why this was so hard to communicate via the podcast. But, kudos to April for trying to make that happen.

I opened and closed the discussion by talking about compromise.  Right now the debate on hatcheries is for all or nothing on either side, at least in most cases. I, and we at TU, see a huge swath of middle ground where few people are operating.  We think this is where progress can be made.

Our solution is what we call a Portfolio Approach. We would designate some rivers as hatchery and other rivers as wild, where there are no releases of hatchery steelhead.  Each decision would need to consider the specifics of a particular watershed, but in general, we would suggest that those rivers that have the capacity to produce enough fish for a CnR fishery are best served by being all wild.  Those watersheds where wild steelhead are struggling and do not produce enough fish for a CnR fishery are better candidates for hatcheries.  This is generally consistent with our steelhead management plan in Washington state, though we seek to have whole watersheds set aside for wild fish and fisheries rather than just parts of watersheds.  In essence it helps achieve recovery goals for wild fish and promotes fishing opportunity, both for kill of hatchery fish and CnR of wild fish.

The benefits of the approach are numerous. It helps achieve recovery goals by securing the strongest populations as wild strongholds without hatchery effects, and provides CnR opportunity on wild steelhead. At the same time, it provides harvest opportunity for those anglers that prefer hatchery steelhead.  It also ensures that we can do large-scale experiments that are needed to answer some of the tough questions about how to best manage hatchery fish and the extent to which wild fish can recover in absence of hatchery effects.  Given the contentious tone of the debates, the latter seems particularly important if we are ever to find a better agreement on the hatchery v. wild issue.”

~John McMillan

Posted in Blog