Are We Making Fish Stupid?

How do you track search image development? You watch fish feed from little homemade feeders, of course.

I’ll start by acknowledging that this is a bit of a bold idea, but bear with me.
 
Researchers recently determined that egg incubation temperature decreases the social learning ability of adult lizards. They conducted an experiment where they incubated some eggs at about 80°F and others at 86°F. Once those lizards reached adulthood, they tested their ability to socially learning a new behavior. Specifically, lizards were allowed to watch videos of other lizards opening a sliding door (a behavior that lizards don’t usually know how to do). After watching the videos, the researchers tested whether lizards hard learned to open the sliding door for themselves. Lizards that were incubated at warmer temperatures were significantly less capable of socially learning the behavior, and thus were less successful at opening the door compared to lizards incubated at the cooler temperatures.
 
So what? Why does a trout biologist care about lizard egg incubation, social learning, and sliding doors? Okay, maybe I don’t care about sliding doors. But, social learning, or simply learning by way of watching or imitation, is something that humans take for granted. We watch a friend solve a puzzle a certain way and we instantly know how to solve the puzzle the same way. Or, if we see a group people heading for a different register at the store, we’re inclined to follow thinking they found a faster way to check-out. Humans use social learning all the time, both consciously (like solving a puzzle) and unconsciously (like finding a new cash register). But, what about fish?
 
My first few research projects as an undergraduate all focused on understanding how trout acquire important information about their environment. The underlying assumption is that every individual learns information the hard way via trial and error where each fish has to learn everything on it’s own. While some information is acquired this way, it’s not a very good learning strategy for most things. It can take a long time to develop a new behavior or learn about a threat, and a fish only gets one chance to learn about a new predator before it’s eaten. So, one alternate strategy is to pay close attention to the behavior of other fish, and pick up new information via social learning.
 
Social learning speeds the learning process, and is particularly helpful in situations where information changes quickly and individuals need to be ready to adapt to their surroundings. For example, and completely hypothetical, streams are subject to rapid changes in flow conditions, which can also change where the best location is for a trout to sit. A fish can try to roam around and actively determine where to go as flow changes. But, the individual is unlikely to gather all the information before flow changes again, plus moving around exposes them to a lot of predation. Alternatively, a fish can use social learning to watch to see what all the other fish are doing around them, and use that information to update their map of the stream without really moving.  
 
As it turns out, trout are incredible social learners. One of my first research projects focused on determining how brook trout gather information about changing food resources in a stream.  As many anglers know, prey availability changes frequently in small streams, and trout have to constantly make decisions about whether something floating past them is food, a stick, or potentially something lethal. So, they typically develop a search image for a few insects, and let everything else float past (this is why it’s important to “match the hatch” and pick flies that resemble bugs currently in the stream). When food resources for which a fish has developed a search image to run out, trout have to learn a new search image and target a type of insect. It turns out, it can take over two weeks for a fish to develop a new search image on its own.  That’s two weeks a fish could go with very limited food as it tries to learn what to eat. But, if a fish is watching another fish eat a new type of insect, it will develop a search image almost instantly. Here, social learning leads to more calories that can be used for growth, reproduction, and even survival.  
 
Another project I did showed that brook trout also use social learning to avoid interacting with other fish they know will outcompete them. Brook trout readily fight with one another for spots in the stream that have the best access to food, concealment from predators, and where flow is not too fast or slow. Naturally, the most aggressive fish (which is usually the biggest) has access to the best spot, number two has the second best spot, and so on down the chain. There’s a benefit to occupying spots of highest quality, but it’s dangerous for a fish to pick a battle with another fish that it is going to lose to.
 
So, how does a trout decide who to fight? As before, it could be a trial and error process wherein a fish fights with most of the other fish around it to determine its rank. But, fighting is energetically costly and can be lethal, so trout try to avoid interactions when possible. To do that, trout watch other fish compete, and from those observations learn which individuals are more and less dominate. It’s like watching a series of playground fights to identify the bully you never want to mess with. By watching other fish compete, a fish can socially learn the competitive ability of many individuals in a pool without ever having to interact with them directly.
 
So, brook trout use social learning to find new food resources (which increases energy that can be used for growth and reproduction) and to limit competitive interactions with rivals (which decreases energetic output and the chances of injury). If increased stream temperature during incubation decreases social learning as it did with the lizards, what effect will that have on trout?  It’s hard to say. We know that trout use social learning, and we can speculate on the energetic benefits to using social learning. But, we don’t know exactly what happens if trout suddenly lose the ability to learn socially. Could we see reduced growth, reproduction, competitive ability, and extirpation?  Yes. But, it would be hard to say that those outcomes happened because of a loss in learning rather than the effects of some other stressor such as stream temperature rise or competition with nonnative species. It’s difficult, perhaps impossible, to isolate all of those stressors from one another.  
 
But, what this highlights is that climate change is more pervasive that we probably think about on a daily basis.  Yes, stream temperature rise can make certain streams too hot for trout to occupy.  But, there are negative consequence of stream temperature rise that occur before population extirpation and that may affect more subtle aspects of fisheries ecology and behavior. 

---