Every year, people haul 1 to 3 trillion fish from the ocean. There’s little doubt that this is too many — fish populations have collapsed worldwide. But there’s another thing we are doing to fish: By killing them in such large numbers, we are exerting enormous evolutionary selection pressure. The adaptive changes that result could increase the chance that fish species will survive, but may pose a problem for the people who depend on them.

By preferentially catching the kind of fish we like — usually big ones — we are inevitably creating a disadvantageous situation for such fish. They’re less likely to survive and reproduce, which reduces the presence of their genes in the next generation. We are, in effect, creating evolutionary pressure for fish to be small. In that way, policies dictating that only large fish should be caught, and small ones left alone, could be counterproductive.

Mikko Heino, an evolutionary biologist at the University of Bergen in Norway, has spent more than 20 years trying to document and understand how humans are changing the way that fish evolve. Because it’s very hard to directly demonstrate genetic changes due to fishing, he has used a diversity of other approaches, from lab experiments to historical data analysis to computer modeling of long-term evolution.

As Heino and two coauthors explain in the Annual Review of Ecology, Evolution, and Systematics, there probably is no way for us to catch fish without evolutionary side effects. Yet he believes that taking evolution into account might put an end to well-intended but ill-advised policies and help develop more sustainable ways to manage fisheries.

This interview has been edited for length and clarity.

Overfishing is a global issue today, and it seems impossible to inventory all populations to know what exactly is going on. So how do you go about studying the impacts of fisheries?

First of all, we have learned a lot from studying fisheries where there is a tradition of data collection. In Norway, for example, these data go back almost 90 years, which reflects the very high importance of some of these populations to local economies. Because those catches were naturally fluctuating, there was a desire to better understand where these fluctuations were coming from, because they had such a huge impact on people’s livelihoods.

By using these kinds of data, we were able to show, for example, that a dramatic collapse of the cod population off the coast of Canada in the late 1980s and early 1990s was preceded by a decrease in the average size of individuals, and earlier reproductive maturation. Follow-up studies have found that similar patterns are affecting dozens of other commercially important fish.

To really demonstrate evolution, one has to show genetic change. Have you been able to?

Not yet. All our evidence so far is based on changes in the appearance of the fish, not their genes. I have been involved in a number of projects attempting to show genetic change, but it turns out to be challenging. This partly reflects the highly polygenic nature of traits such as body size and maturation. Almost every gene has an effect on body size, and so does the environment. The selection response is therefore quite spread out across the genome, and not very large at any single gene. Yet the fact that these changes are so common does suggest to me that adaptive evolution is the most likely explanation.

Another way we are approaching this is by doing experiments in the lab, often using smaller species such as guppies, to show what the evolutionary effects may be when we regularly remove the largest individuals. These experiments are necessarily a simplification of real life, but they may help to focus our attention on aspects we may not have thought about, such as the importance of cannibalism when different generations live in the same place, which is the case in certain populations of cod, for example.

Last but not least, we can feed the historical and experimental data into computer models to extrapolate our findings into the future and find out what might happen after many generations, or at much larger scales. These are things that would otherwise be too expensive or time-consuming to investigate, and modeling them may help us understand patterns we observe in real life, or even inform new policies and regulations.

Photograph of three groups of fish of different sizes. A penny is shown for size reference.

In 2002, an experiment at SUNY Stony Brook found that selectively removing the largest or smallest fish from experimental populations of Atlantic silverside (Menidia menidia) could create a twofold difference in body weight after four generations. (Left: fish from the original population. Right top: population in which small individuals were removed. Right bottom: population in which large individuals were removed.) In 2019, a team including the original authors did a genetic analysis of some of the fish, which had been stored in a freezer after the experiment. They found that many growth-related genes were involved in the changes, but that the precise nature of the changes differed from one experiment to the next. In some cases, whole blocks of genes changed together, enabling a rapid response to the selective pressure exerted by the experiment. The findings mimic the kinds of effects that selective fishing could create and show that there’s more than one genetic way for fish to evolve to get smaller or larger.


Usually, when human activities are a threat to plants or animals, there is a sense of relief when scientists discover they are evolving in response. So why would it be bad news if fish do the same?

Well, it depends whether you take the perspective of the fish, or of humans using them as a resource.

For the fish, it is probably a good thing, as it will make them more resilient to fishing. And from a larger sustainability perspective, that is also good for us. But if you are interested in maintaining the same levels of yield, the problem is that fish are likely to evolve in a way that lowers fisheries’ productivity. So populations and species may still be around, but they will become types that are less valuable for us.


As human populations keep increasing, this is a problem for food security, perhaps not in Western economies, but in some developing ones, and in many local communities that depend on fisheries.

That being said, fisheries can also lead fish into a kind of evolutionary trap. Evolution is myopic — it can only respond to changes in the environment, not predict them, so it will reflect the characteristics that were beneficial in the recent past, with no guarantees for the future. But when we push fish to evolve a smaller body size, this may, for example, decrease their physiological ability to make the long spawning migrations that species such as cod and tuna depend on.

In the review, you point out that even if we don’t fish selectively, so that every individual fish has the same chance of being caught, we are still causing evolution. How so? Is there really no way to avoid this?

It is essentially unavoidable. When we fish, we impact the circumstances for the population, because we are removing individuals and increasing the rate of mortality. Even if the relative risk of dying is exactly the same for all the fish in the population, irrespective of their traits, if you elevate overall risk — which we always do when we fish — it will cause selection for fish to live faster. This means they will be maturing and reproducing at earlier ages and at smaller sizes, because the ones that wait too long may not get to reproduce at all.

The only way to minimize selection would be for our fisheries to exactly replace the impact of natural predators, by removing or outcompeting them, as we are arguably doing to some extent. But from a conservation point of view, eliminating large marine predators is also problematic, of course.

That is not very encouraging. Is there nothing else we can do to minimize our impact?

The most robust advice I can give is that when fishing for single species, aiming for intermediate-sized fish, or at least not singling out the large ones, is probably the best approach to reduce selection.

Photograph of a shirtless man in a captain’s hat standing next to a woman in a bathing suit and a little girl holding a pet dog. A haul of large fish is displayed on the ground in front of them and strung up behind them.

Captain Tony Tarracino and his wife and daughter show off a catch of goliath grouper in 1958, in the Florida Keys. Big fish have been getting smaller ever since. A 2009 study found that the average size of trophy fish displayed in pictures like this was less than half what it was in the late 1950s.


That is almost the opposite of what many fishery practices and regulations currently prescribe, which is to release or avoid small fish and only keep the big ones. Why do these rules exist?

The classical thinking in fisheries biology goes that we should catch fish when they reach their most productive size, while leaving alive the ones that are still going to grow. Conveniently, this is also fully compatible with the Western desire to eat larger fish. But it is not as sustainable as we like to believe.

Does this mean that more recent regulations requiring fishing vessels to land everything they catch, even if it can’t be sold — which many fishermen consider to be wasteful — are indeed a good idea?

They are a move in the right direction, because it makes some selection pressures less strong. But it would be naive to believe that this solves the whole problem. Like I said, a large part of these selective pressures is the result of shortening average lifespans, which is caused by the increased mortality risk. This will still occur if you are harvesting smaller individuals as well, unless you dramatically reduce the harvest of large ones. So lowering the absolute levels of harvesting will also be important.

This is desirable for many reasons, of course, even without considering the evolutionary perspective. If fish populations are large, you can get a larger catch for the same investment of effort, fuel and salary. So there are good arguments for fisheries management aiming for more moderate levels of exploitation. The exact level depends on how we weigh different costs and benefits. But some of our data and models can provide valuable input to that discussion.

Human fishing affects the size of the fish population, and over generations, the size of the fish. Scientists have found that when fisheries go after mostly large fish, they create evolutionary pressure that favors the survival of smaller fish. Eventually, intense pressure from fishing could lead to food supply issues. In this video, Mikko Heino, a biologist at the University of Bergen, speaks with Knowable Magazine to explain how this effect works and what we can do to minimize human impacts on fish evolution.


Are there any efforts to try to modify those models so they can be used to fine-tune policies?

It is something we have been advocating, but many of the fisheries biologists who advise these policies are more concerned with the more immediate effects of overexploitation. And policies are often slow to change, so including these evolutionary considerations will probably be a long-term process. In the meantime, we can use these models to create more awareness, also among the public.

Is there anything consumers can do? Has your research changed the kind of fish you buy?

Myself, I try to eat more smaller fish like herring, sardines or anchovies, and fewer top predators such as cod, which have to consume many of those smaller fish to grow to the size at which they are usually caught.

From the global food production perspective, I think this makes a big difference. In some cultures, people eat a lot of tiny fish, whose mortality is naturally high and therefore less problematic. But in the West, many of us are not used to eating small fish. It is partly just a matter of informing people.