Related Resources
Lion Sleep
Links to Sleep Research from the Center for Sleep Research, Siegel Lab
| Join FONZ to receive ZooGoer in your mailbox! |
To Sleep, Perchance to Dream
By Jessica Marshall
Oh sleep! it is a gentle thing,
Beloved from pole to pole.
—Samuel Taylor Coleridge
Anyone who has ever stayed up too late and regretted it the next day knows just how much humans need sleep. Animals need it, too. Every animal studied so far—from whales to octopuses to fruit flies—sleeps, although animal sleep takes various forms, and even among mammals the human eight hours is not the norm. Horses, elephants, and giraffes, for instance, sleep only about two to four hours a day, while bats and opossums sleep up to 20.
 |
| All land mammals, including this giant panda (Ailuropoda melanoleuca) at the Smithsonian's National Zoo, experience rapid-eye movement (REM) while sleeping. (Jessie Cohen/NZP) |
It is obvious to us what it means to be asleep, but how do scientists know whether an animal that looks asleep is doing the same thing that we are? In some cases, they can record the animals' brain waves and compare them to those of sleeping humans. But in others, scientists have settled on a few observable criteria: Is the animal in a characteristic sleep posture? Is it unresponsive? Immobile? If it is deprived of sleep, will it then sleep longer to make up for lost zzz's?
Another behavioral criterion for sleep is that it is rapidly reversible, unlike hibernation, which takes a long time to enter and emerge from. Jerome Siegel, a neuroscientist at the University of California Los Angeles, says that hibernation is very much like sleep: Both help animals to conserve energy, but hibernation is more extreme. In hibernation, body temperature drops—sometimes to just a few degrees above freezing—and the brain truly shuts down, except for low-level brain activity needed to maintain functions such as breathing and circadian rhythms. Less drastic than hibernation is "shallow" torpor, in which an animal's body temperature drops, but not as low and for a shorter period of time. Some animals—like hummingbirds and hamsters—can enter and emerge from torpor daily.
"[Torpor] may be a compromise," says neuroethologist Niels Rattenborg of the Max Planck Institute for Ornithology in Starnberg, Germany. Animals in torpor save more energy than in sleep, but they are less responsive to potential threats. However, work by Craig Heller, a professor of biology at Stanford University, suggests that hibernation and torpor are not a substitute for sleep. He found that hibernating golden-mantled ground squirrels (Spermophilus lateralis) raise their body temperatures every five to seven days for about a day. During that time, they undergo slow-wave sleep—the deepest stage of sleep, characterized by slow brain waves—that is deeper than any other sleep they have when not in hibernation. Then they cool back down. So, even when animals have been inactive and resting during hibernation, they seem to need sleep.
"Current thinking is that brain structure is lost during a bout of deep hibernation, and it is growing back during these periods between bouts of hibernation," Heller says. "Until all of these neural connections are restored, the animals are less arousable and therefore sleep deeply." Animals emerging from shallow torpor also go straight into slow-wave sleep, which suggests a need for sleep on top of the time spent in torpor.
REM Sleep
One of the most inexplicable elements of sleep is REM, or rapid-eye movement, named for the fluttering motion of the eyes under the sleeper's eyelids. Unlike in deep sleep, which is characterized by slow brain waves, the brain during REM sleep is just as active as when it is awake. Dreaming happens during REM sleep in humans.
All land mammals and birds have REM sleep, but reptiles, amphibians, and fish appear not to. Rattenborg suggests that REM sleep may be involved in intelligence (as slow-wave sleep may be), because both birds and mammals have high cognitive abilities. "We have always joked and used the term ‘birdbrain' to indicate that somebody's stupid," he says, "but birds are able to perform cognitive tasks that exceed some mammals."
If REM sleep is a hallmark of intelligence, then Stephen Duntley, director of the Sleep Medicine Center at Washington University in Saint Louis, wonders whether another notably smart class of animals—the cephalopods, including octopuses, squid, and cuttlefish—also experiences REM sleep. "They are incredibly intelligent," says Duntley, who has studied sleeping cuttlefish.
 |
| Sleep research on invertebrates is not extensive, but cephalopods like this common cuttlefish (Sepia officinalis) may have two-phase sleep similar to that of birds and mammals. (Mehgan Murphy/NZP) |
Like all cephalopods, cuttlefish have tiny organs called chromatophores in their skin that they use to change color in response to their environment. Duntley hypothesized that if cuttlefish "dream," they would continue to flash colors with their chromatophores, because they sleep in a protected environment where their color changes would not make them visible to predators. He found that pharaoh cuttlefish (Sepia pharaonis), which sleep under the sand, begin sleep with a light, monochromatic appearance, but every ten or 15 minutes they flash a geometric pattern of darker color unlike anything they display while awake. This suggests that cuttlefish also have alternating two-phase sleep, like the slow-wave and REM sleep of birds and mammals. It is still only a hint, though. Duntley has yet to monitor cuttlefish's brains during these flashing periods to determine whether this is truly similar to REM sleep.
A handful of mammals throw a wrench in the theory that brainy animals need REM sleep. Dolphins and other marine mammals, for instance, are some of the most intelligent animals known, yet "there have been no published reports documenting REM sleep in cetaceans, making them the only studied mammals in which this state has not been observed," writes Siegel in a 2005 paper in Nature. Further, most other newborn mammals get more REM sleep and spend more time sleeping than adults of their species, which suggests that sleep is important for brain development, but Siegel has found that dolphin neonates do not have any extended periods of sleep.
Another anomaly in REM sleep is the platypus (Ornithorhynchus anatinus), one of the most primitive mammals. The platypus does not tip any scales for cognition but has more REM sleep than any other mammal. Platypus REM sleep is "spectacular," Siegel says, with dramatic twitching of the bill and legs. Siegel's work with the platypus and one of the few other egg-laying mammals, the short-beaked echidna (Tachyglossus aculeatus), adds to the conundrum: His measurements of these animals show that they have REM sleep in the brain stem—a primitive part of the brain—and not the neofrontal cortex, which is more developed in higher mammals. "REM sleep is quite a mystery," he concludes.
To Each Its Own Sleep
Sleep might be universal, but animals have evolved different sleep patterns that suit their lifestyles. As omnivores who cannot seek food in the dark (24-hour diners notwithstanding), humans typically get their roughly eight hours of sleep in blocks, mostly during the night. But rats sleep in spurts that can happen any time of day, and they sleep ten to 12 hours a day.
Birds have varying sleep habits, too. As you might guess from the rooster's famous crowing at dawn, many birds sleep in one stretch, usually at night. But shorebirds sleep during the day or night, because they rely on low tides for finding food, which can happen at any time of the day. These birds have sensors in their beaks that allow them to find food even in the dark. Rattenborg and his colleagues studied white-crowned sparrows (Zonotrichia leucophrys gambelii) in the lab and found that they reduced their time spent sleeping by two-thirds during their migratory season, even though they stayed in their cages, hopping around and flapping their wings. Despite this lack of sleep, the birds showed no sign of sleep deprivation: Their performance on cognitive tasks was unchanged, and they did not sleep any extra to catch up after the migratory season ended. Now Rattenborg plans to track the birds while they are flying, because "It is possible that when they are up in the night sky with nothing to run into, they might nap," he says. The common swift (Apus apus), for example, spends almost its entire life in flight and may sleep on the wing. To test this, Rattenborg is using a wind tunnel specifically built for birds that has very smooth air flow. He is also using a mobile device to track the brain-wave patterns of birds in flight.
In general, herbivores sleep less than omnivores, which sleep less than carnivores, perhaps because herbivores need to spend more of their waking hours eating to meet their food needs. Small herbivores tend to sleep longer than large ones, perhaps because they need to conserve more energy to maintain their body temperature.
 |
| Most animals sleep where they they are relatively safe from predators. This red panda (Ailurus fulgens) is sleeping in a tree on the Zoo's Asia Trail. (Mehgan Murphy/NZP) |
Herbivores may also spend less time snoozing because they can't sleep as safely as large predators. This may be true for certain animals that can sleep only in exposed environments, Siegel says, but he is not convinced that sleep poses a predation risk in most cases. "People might think, in theory, that animals are more vulnerable during sleep," Siegel says. "I think that is a fallacy." Most animals have evolved ways of sleeping that involve decreased risk, such as sleeping in burrows where they are relatively safe, he says. If sleep were more dangerous than being awake, animals might have evolved ways to rest parts of their brain at a time, rather than conking out all at once. But Rattenborg and ecologist Steven Lima of Indiana State University in Terre Haute propose that in most cases, for animals with complex, highly networked brains such as birds and mammals, unresponsive, "blackout" sleep may be safer than if these animals turned off pieces of their brain and wandered around without their full wits about them.
Yet some birds and mammals do forgo blackout sleep; instead, they keep one eye open and half their brain awake. For example, when mallard ducks (Anas platyrhynchos) sleep in groups, those on the outside edge sleep with their outside eye open and the opposite side of the brain (which controls the outside eye) awake. Presumably, the birds on the edge have to keep an eye out for predators while the birds in the middle can go fully to sleep. This behavior clearly suggests that birds are vulnerable during sleep, Rattenborg says.
Bottlenose dolphins (Tursiops truncatus) also sleep by closing one eye and shutting down the opposite half of their brain. While in this state, they may remain still or they may swim, avoid obstacles, and move both sides of their body, violating the "immobility" criterion for sleep. This suggests to Siegel that under such conditions the "sleeping" half of a dolphin's brain is not actually asleep, or is at least not experiencing blackout sleep.
But southern fur seals (Arctocephalus spp.) in the water truly sleep with half their brain, Siegel says. When one half of their brain goes to sleep, their opposing fins and whiskers become paralyzed. "When they are in water, which is where they live for six months of the year, they have very little REM sleep," he says. But when seals come on land, they sleep like dogs: Their whole brain goes to sleep and the brain-wave patterns match typical land-mammal patterns. They don't seem to need to catch up on the missed REM sleep.
Reptiles, amphibians, and fish all show sleep behavior. Although few studies have characterized their sleep, animals in these groups do not appear to have either slow-wave or REM sleep. Reptiles show some patterns that are "slow-wave like," Rattenborg says, but this is in strong contrast to the slow-wave patterns of birds and mammals, which are very easy to detect. Because reptile studies are not consistent, Rattenborg says scientists really don't know whether reptiles undergo slow-wave or REM sleep.
Work with insects and cephalopods has shown that invertebrates sleep, too. Paul Shaw of Washington University in Saint Louis reported in 2000 in Science that fruit flies (Drosophila melanogaster)sleep, based on behavioral measures: They don't respond if researchers tap lightly on their containers or pass a shadow over them. If the scientists keep the flies awake, the flies stay immobile for longer periods afterward, apparently to catch up on missed sleep. When Shaw's team gave the flies caffeine, they slept less.
Even some jellyfish may sleep. Jamie Seymour, director of the Tropical Australian Stinger Research Unit at James Cook University in Cairns, Australia, began tracking individuals from one species of box jellyfish (Chironex fleckeri) by gluing radio trackers to them. At first, Seymour and his colleagues thought the tracking devices were falling off, because the signals were falling to the ocean bottom. "It wasn't until we got in the water that we found the jellyfish were actually there, they just were not moving," he says. "Around about three or four o'clock in the afternoon, they would drop out of the water column and sit on the sea floor until the sun came up the next morning. The animals stopped pulsing. The tentacles were completely relaxed." Among jellyfish, C. fleckeri are unusual: They are active predators that can propel themselves through the water, and can grow two to three millimeters a day throughout most of their life span of a few months. They even show daily growth rings that suggest a 24-hour cycle marked by fast and slow growth. Seymour concluded that the jellyfish use their time on the bottom to redirect energy from moving to growing. "Whether it's sleep or not, I don't know," he says.
What Is the Purpose of Sleep?
Even though sleep may be universal, scientists are divided on why animals need it. Rats and flies deprived of sleep die faster than if they are deprived of food. This suggests that at least for some animals, sleep is essential.
 |
| Birds, like this American flamingo (Phoenicopterus ruber), exhibit interesting sleep behaviors. Some species seem to sleep with one eye open; others may sleep on the wing. (Mehgan Murphy/NZP) |
Evidence is growing that animal brains need sleep for learning and consolidating memory. "The data are getting stronger each year," Rattenborg says. Studies of rats have shown that they experience more REM sleep after periods of learning, and that if deprived of REM sleep, they perform worse in mazes. Humans perform newly learned tasks faster after sleeping, but show no improvement after the same amount of time if they haven't slept. Shaw showed recently that the same phenomenon happens in fruit flies. Female fruit flies that have already mated reject males, but unmated females readily accept male advances. Shaw "taught" males that females would not accept their advances by introducing them exclusively to mated females. After this training, some males were sleep deprived for four hours while others were allowed to sleep as they desired. The males allowed to sleep slept extra long—more than if they had been kept awake all night. Three days later, Shaw exposed both sets of males to unmated females. Those that had slept did not court the females, indicating they had learned that females were unreceptive. Sleep-deprived males, however, showed no such restraint in courting the willing females.
Additional evidence from studies by neuroscientist Matthew Wilson at the Massachusetts Institute of Technology in Cambridge suggests that the brain replays events from the previous day during sleep, supporting the idea that the brain consolidates memories or learns during sleep. Wilson monitored the brains of rats as they ran a maze, measuring which neurons were active. During sleep that followed, he detected the same patterns of activity in the same sets of neurons that were activated when the rats were running the maze.
But Siegel finds much of this evidence unconvincing. He thinks the answer to why animals sleep is much more fundamental—they have evolved to sleep when there is no point in staying awake, when they are better off staying out of harm's way and saving their energy. For example, most cats sleep 12 to 15 hours a day. For a lion that has just killed an antelope, it makes sense to rest and digest for the next several days rather than using up energy or risking being injured when it does not need food, Siegel says.
Two species of bat (the big brown bat and little brown myotis, Eptesicus fuscus and Myotis lucifugus) sleep most of the time, because their food is available only for a few hours at night when their particular diet of insects emerges. They are better off sleeping the rest of the time, he suggests. In humans, the brain uses 25 percent of our energy at rest, so shutting it down could save significant energy, he says, and even small energy savings are significant in evolutionary terms.
Siegel acknowledges that for some animals sleep serves other functions, but he believes that these functions moved into sleep under evolutionary pressures that arose after the primary function of energy conservation. In humans, for instance, a growth hormone is released only at night in very deep sleep, and children deprived of such sleep by stress show stunted growth. But in dogs, a similar growth hormone is released during the day. "There must be a reason for this in humans," Siegel says. "There may be a whole lot of things like that that have migrated into sleep in different species that aren't universal."
Animal Dreams
If Siegel is right, certainly one such phenomenon is dreaming. Dreams are a fascinating aspect of human sleep, but how widespread is dreaming in other animals? You have probably seen a dog twitching in its sleep; most people would assume it is dreaming. "From an evolutionary perspective, I think it's very unlikely that dreaming emerged for the first time in humans," Rattenborg says. Cats whose brains have been damaged in the part that normally keeps them paralyzed during sleep stand up and seem to attack imaginary things in their sleep: It appears they are acting out their dreams. The fact that all land mammals have REM sleep, which is associated with dreaming in humans, suggests that dreams are widespread. However, dream reports are uncommon or very simple in children between three and five years old, and people with certain types of brain damage have normal-looking REM sleep but never report dreams. So animals may not dream at all the way that we do, and because we cannot ask animals about their dreams, we will probably never know.
Nonetheless, researchers believe we have much to learn by studying animal sleep. Shaw's work with fruit flies has recently uncovered a molecule—the enzyme amylase—in their saliva that accumulates when they are sleepy. His team then found the same molecule in the saliva of sleep-deprived humans. This finding could eventually lead to a measure analogous to blood alcohol content to determine if bus drivers or shift workers are dangerously low on sleep. Manipulating the genes and monitoring various areas of the brain involved in sleep—experiments that can easily be done on fruit flies and mice—are also providing insights into what drives sleep at the molecular level. With many species' sleep never tested, the evolutionary patterns of sleep behaviors remain unclear. So sleep researchers can rest easy. There will be plenty of questions to keep them awake at night for years to come.
—Jessica Marshall is a science writer living and sleeping in St. Paul, Minnesota.
ZooGoer 36(6) 2007. Copyright 2007 Friends of the National Zoo.
All rights reserved.
