Sleep-deprived jumping spiders spent 80% more time in REM-like sleep

A jumping spider kept awake by fruit flies for one night spent the following night catching up, and almost all of the extra time went into a twitching, eye-flickering phase that looks like REM sleep. Across thirteen fringed jumping spiders, time in that REM-like state rose by roughly four-fifths after a single disrupted night.

That rebound is the strongest sign yet that spiders do more than sit still in the dark. Animals that need to repay lost sleep are regulating it, which is one of the defining marks of true sleep in mammals and birds. The authors say this is, to their knowledge, the first time REM rebound has been shown in an invertebrate.

The spider is Portia fimbriata, a small hunter from northern Australia that is about the size of a one-euro coin when fully grown. Portia species eat other spiders, and they are known for cunning: they pluck the silk of a web to imitate a struggling insect, then ambush the web’s owner when it comes to collect.

The study, led by Nadja Geiger and Daniela Rößler of the University of Bonn with colleagues in Germany, Italy, Australia and the United States, shows that the spider’s night is organised into at least two distinct stages. A quiet phase comes first. Bouts of twitching legs, trembling mouthparts and moving retinas interrupt it again and again, and the spiders are hardest to wake during those bouts.

How they did it

Sleep is hard to prove in an animal whose brain you cannot record, so the researchers relied on the behavioural tests that sleep scientists use for any species: a typical resting posture, a predictable daily rhythm, reduced response to the outside world, and compensation after sleep is lost.

The spiders lived in plastic dishes 12 centimetres across and were filmed for several days and nights, with infrared cameras after dark. Their activity was tracked for 60 hours under different lighting regimes. They were active mainly during the light phase, and a roughly daily rhythm persisted in constant near-darkness, a sign of an internal clock rather than a simple reaction to light. Under constant light, rest became fragmented.

Each night followed a similar script. The spiders looked for a sleeping spot, groomed, and hung upside down beneath a small silk platform, most of them within half an hour of the lights going off. Over the night their bodies slowly drooped and the small appendages beside the mouth, the pedipalps, drew inwards. In ten spiders filmed in detail, the nightly sleep window averaged about 10.7 hours, including short waking spells.

In those same ten spiders, the team counted 109 REM-like episodes, each lasting about 83 seconds on average. Jumping spiders cannot move their large front eyes, but they can move the retinas behind the lenses, and during these episodes the retinas darted back and forth while legs curled and muscles twitched. The first episode came about four hours after sleep onset on average, but the timing ranged from just over an hour to about ten hours.

To test how deeply the spiders slept, the researchers played a standard sound and measured how long each spider took to react. Awake spiders responded in about 1.7 seconds, spiders in the quiet phase in about 2.7 seconds, and spiders in a REM-like episode in about 4.3 seconds. Four spiders in REM-like episodes did not respond within the ten-second window at all, yet reacted immediately when tested again while awake, which argues against simple deafness or exhaustion.

The final test was deprivation. For one night, the team released 25 live fruit flies into each enclosure, a disturbance a hunting spider cannot ignore. On the recovery night, the overall sleep window grew by about 8.5 per cent, but REM-like time jumped from an average of 9.4 minutes to 16.9 minutes. The spiders had more episodes, not longer ones. Five undisturbed control spiders showed no such increase over consecutive nights.

Why a sleeping spider matters

Spiders split from the lineage that led to vertebrates more than 500 million years ago, and their nervous system is built very differently from ours. Finding a sleep architecture that resembles the vertebrate pattern in such a distant animal suggests that sleep stages are either very old or a solution that evolution reaches more than once.

“If animals with such different brains and sensory worlds show structured sleep with REM-like phases, it suggests that the basic architecture of sleep may be far older and more widespread than we thought,” Rößler told ABC News in Australia.

The comparison now extends across three distant branches of the animal tree. Octopuses and cuttlefish show an active sleep stage with flickering skin patterns, vertebrates have REM sleep, and now a jumping spider shows a twitching stage that the body tries to recover when it is lost. “Seeing that spiders show a clear REM-like sleep pressure, essentially a need to compensate for lost REM-like sleep, strongly suggests that this stage fulfills an important biological function rather than being an incidental feature of rest,” Geiger and Rößler told ScienceAlert.

The work builds on the same group’s earlier study of another jumping spider, Evarcha arcuata, in which young spiders hanging at night showed retinal movements and leg curling. That study could show the twitches but could not show that the spiders were actually asleep. The new paper fills that gap, which is why Thomas White, an entomologist at the University of Sydney who was not involved, called it “far and away the most complete behavioural case for sleep in any arachnid”.

What it doesn’t settle

The biggest limit is the brain. No one recorded neural activity, so the study cannot say whether the spider’s REM-like stage involves anything like the brain activity of human REM sleep, and it cannot identify a spider counterpart to deep non-REM sleep. The resemblance is behavioural, and the authors are careful to call the state REM-like rather than REM.

Dreaming is further away still. Twitching limbs and moving eyes say nothing about whether an animal experiences anything while asleep. The team suspects the REM-like stage helps Portia process what it learned while hunting, and raises the possibility that the twitches are a form of behavioural replay of daytime actions. That idea has not been tested.

The samples were small: ten spiders for the detailed night recordings and thirteen for the deprivation test, all in a laboratory. The difference in reaction time between the quiet phase and REM-like episodes pointed in the expected direction but did not reach statistical significance after correction, and the quiet phase itself mixed responses that resembled wakefulness with others that resembled deeper sleep. Where quiet wakefulness ends and light sleep begins remains blurry.

It is also one species. Portia is unusually visual and unusually clever for a spider, and the results may not apply to web-building spiders that hunt by touch.

Common questions about spider sleep

Do spiders sleep?

At least one species clearly does. Portia fimbriata meets the standard behavioural tests for sleep: a fixed resting posture, a daily rhythm driven by an internal clock, slower reactions while resting, and extra sleep after a disrupted night.

Do spiders dream?

Nobody knows. Jumping spiders show a REM-like stage with eye and leg movements, but without brain recordings there is no way to tell whether they experience anything. The researchers think the stage may help with learning and memory, which is not the same as dreaming.

What is REM rebound?

When people or other mammals lose REM sleep, they spend more time in REM on the following nights to make up for it. The spiders did the same after one disrupted night, raising their REM-like time by about 80 per cent, which shows the stage is regulated rather than random.

Why do jumping spiders twitch at night?

During REM-like episodes, the retinas inside their main eyes swing back and forth while their legs curl and muscles twitch. These bouts last about a minute and a half and return several times a night.

What comes next

The paper appeared online in the Proceedings of the National Academy of Sciences on 26 October 2026, with work funded by the German Research Foundation’s Emmy Noether Programme. The team is now comparing sleep across the wider spider family tree and is starting fieldwork in northern Queensland and Western Australia to see how spiders sleep in the wild, including how artificial light at night disturbs it. Further studies aim to record what happens in spider and insect brains during sleep and to test whether REM-like sleep is linked to the learning that makes Portia such an effective hunter.

Reference: Geiger et al., “Behavioral evidence for two-stage sleep in spiders,” Proceedings of the National Academy of Sciences, 2026. DOI: 10.1073/pnas.2618030123

Tags: biology, sleep, jumping spiders, Portia fimbriata, REM sleep, spiders