Read our COVID-19 research and news.

Bruce Kennedy, then an undergraduate student in Garet Lahvis’s lab, video recording ground squirrels in Wisconsin.

Garet Lahvis

Swapping a cage for a barn: Can lab animals be studied in the wild?

The environment a laboratory animal lives in can have a dramatic impact on whether it’s a good model for human disease. A mouse that lives in a shoebox-size cage, for example, gets less exercise than its wild relatives, and thus may not be the best model for studying obesity.

Enriched environments with bigger cages and more toys can help, says Garet Lahvis, but the best way to make animals good models is to take them out of the lab—and, in some cases, study them outside in the great wide world. This could be accomplished with cutting-edge electronics and remote sensors, says the behavioral neuroscientist at Oregon Health & Science University in Portland. He’s presenting his proposal today at the Behavior Genetics Association’s annual conference in Boston.

Lahvis chatted with Science about what studying lab animals in the wild could look like, and why some researchers think it won’t happen. This interview has been edited for clarity and length.

Q: Why did you become interested in this idea?

A: Our lab studies social behavior in mice. We’ve shown that mice have the capacity for empathy when they hear other mice getting an electrical shock, and that mice are gregarious—they like to hang out with each other. But we were studying them in these small, relatively sterile cages—not anything like they’d encounter in the wild. About 6 or 7 years ago, I started thinking, “How could it be normal for you to spend your entire life with only three other individuals in a small room? Are the mice we’re looking at really normal?” Once that door opened, I started to think about everything else that could go wrong with lab animal research.

Q: What’s wrong with current housing for lab animals?

A: Animals have to deal with a massive landscape in the wild. They need to hunt for food, avoid predators, seek mates, and deal with complex environmental variables, from rain to shifting light conditions. If you live in a cage, even a cage with lots of toys and companions, everything is the same. You’re not using your brain.

We haven’t had any big breakthroughs for psychiatric diseases; even with cancer drugs, the success rates have been very low. I think a contributing factor is that we’re keeping these animals in such unnatural conditions.

Q: What’s the solution?

A: We need to make the lives of these animals as close as possible to what their lives would be like in the wild. If you’re studying something on a macro level—like how an “autistic” animal responds to its companions—you have to put the animal in a situation where it’s dealing with a ton of social and environmental variables, and those variables need to constantly be changing. You can’t just put a couple of mice in a cage for 4 minutes and ask how long they spend sniffing each other’s butts.

Q: So take the animals out of the lab?

A: One option is to put the animals in an outdoor pen, so they’re dealing with things like foraging and weather that they don’t have to in the lab.

But we could also do some of this work in the wild. For example, [some] mice live in barns. So we could construct a real barn in the wild, one that would have all the challenges and opportunities of real barn living, except perhaps we’d try to make sure the barn never got too hot or too cold, and that the animals couldn’t escape. Then you could introduce, for example, a bunch of mice that were genetically predisposed to breast cancer.

Garet Lahvis thinks scientists should study lab animals in more naturalistic conditions.

Melissa Weddle

Q: But how could you study such free-roaming animals?

A: You could tag them with a radio frequency tag to keep track of them, and implant them with devices that would allow you to measure physiological variables like heart rate. Then you could inject them with a drug and catch them with traps at regular intervals to see if their tumors were shrinking.

Pretty much everything we can do in the lab, we can do in the wild through remote telemetry and other electronics. You could even use camera traps, for example, to see how animals solve puzzle boxes in the wild. Or you could use probes implanted in an animal’s brain to remotely turn genetically modified neurons on and off.

We’ve done this ourselves with ground squirrels. We’ve used video cameras to compare how these animals behave in the wild to how they behave in the lab.

Q: What are the limitations?

A: I wouldn’t modify the genes of animals we’re studying in the wild, in case they escape. Some critics say it’s also a lot more expensive and labor intensive to study lab animals in the wild.

But we also spend a lot of money on infrastructure, cage cleaning, and other things in the lab—and we spend money on experiments year after year that have no relevance to human health. We should spend more money and do it right. Instead, we’re putting monkeys in mop closets.

Q: How have scientists responded to your proposal?

A: Most of them don’t question the science. They realize this is a better way to do things. But they do question the accounting. They think it will cost too much. So the pressure to change has to come from the outside—from an organization like the National Institutes of Health.

Q: What’s the next step?

A: I’ve convened neuroscientists and wildlife biologists in Oregon to explore how use of naturalistic environments can improve research. The U.S. Department of Agriculture, the Department of Fisheries and Wildlife at Oregon State University, and the National Science Foundation have all expressed interest in exploring this.

Q: Have you stopped doing animal research in the lab?

A: I ended my mouse colony and stopped writing grants to study animals in cages a couple of years ago. If this doesn’t work out, [laughs] my days are probably numbered.