At twilight, a moth takes flight in a thistle patch. It distinguishes the purple bristles of a flower from the thorny vegetation using specialized reflectors like those that make a cat’s eyes glow green in the darkness. The flower entices it with nectar and dusts its feathery wings with pollen.

Suddenly, a streetlight buzzes and cracks on overhead. The moth instinctively changes course toward the light. A nearby bat salivates at the swarm of nocturnal insects in their trance-like ascension and makes an easy dinner of the pollen-coated moth. When the birds break into their morning songs — earlier than usual thanks to all the artificial light — the flowers below remain barren.

study published Wednesday in the journal Nature suggests that light pollution, which has already been shown to disrupt the activities of birds and animals, including humans, also interferes with pollination. The study showed for the first time that illuminated flowers were visited by fewer insects, resulting in fewer fruits.

Eric Warrant, who studies nocturnal insects at Lund University in Sweden and was not involved with the study, called this discovery “alarming and a bit depressing at the same time.”

Eva Knop, the first author of the publication, studies the effect of urbanization on ecosystems at the University of Bern in Switzerland. She watched moths flapping around a streetlight and wondered if the distraction might intrude on their nightly rendezvous with flowers.

To find out, she and her collaborators set up mobile streetlamps in seven rural meadows. Sporting night-vision goggles, they visited the meadows and collected insects from flower tops with a net.

“This was a huge undertaking,” said Thomas Davies, who studies the ecological effects of light pollution at Exeter University and was not involved with the study. They had to capture, freeze, pin and analyze thousands of specimens.

Compared with fields where there were no lights, 60 percent fewer nocturnal pollinators were found on illuminated flowers. There were also 30 percent fewer species, suggesting that light pollution decreases the diversity of pollinators.

Watch: Light pollution across the planet (New World Atlas of Artificial Sky Brightness/Nataliya Rybnikova/ArcGlobe)

Ecologists have long suspected that artificial lights at night might interfere with pollination. Now they have evidence. “This lets us put forward recommendations to people that are, this very minute, designing future lighting systems,” said Travis Longcore, an urban ecologist at the University of Southern California.

“Artificial lights are attractive to insects for reasons we don’t fully understand,” Warrant said. It’s thought that they might navigate by moonlight or fly toward light to escape the canopy of a dark forest.

Lights can also affect insect reproduction, for example, by distracting male moths from the pheromones emitted by females. Predators like bats and geckos have been shown to adjust their feeding habits to forage around lights, an added threat to nocturnal insects. Whatever the cause, more lights probably mean less pollinators visit flowers.

To determine whether fewer visits by nocturnal insects meant less pollination, Knop’s group surveyed the number of fruits produced by plants in their experimental meadows. They analyzed 100 cabbage thistle plants. The buds of these thistles before they flower resemble their cousin the artichoke. After pollination, they produce tiny fruits called achenes, which are similar to what we think of as the seeds in a strawberry.

These flowers were common to all sites and a favorite among pollinators, especially nocturnal ones. Thistles from sites with street lamps produced about 10 percent fewer fruits. That means pollination during the day did not make up for the deficit at night caused by the lamps. It could also mean that the lights influence the physiology of the plants directly.

Even plants at the periphery of the glow, 15 to 20 yards from the lamps, produced fewer fruits. Longcore predicts that plants outside of the direct beam of streetlights could also be affected by the “sky glow” that occurs in urban areas.

Streetlights could reduce fruit production in at least two ways, Longcore said: indirectly, by misdirecting pollinators, and directly, by influencing the plants’ sense of time. Like humans (and birds and insects), plants have circadian rhythms. Plants use shifting light spectra as signals for where to grow and when to flower.

He shared a story about a soybean field on the outskirts of Lincoln, Neb., where the plants surrounding streetlights failed to set seeds. These soybeans were waiting for shorter days to signal “winter is coming,” but under the artificial glow, that signal never came. Soybeans, which are self-pollinated, do not rely on insect pollinators, so the effect must have occurred from the lights directly.

Plants and insects also differ in how they respond to certain wavelengths of light. The ratio of red to far-red light tells a plant that it’s in the shade and needs to grow its way toward the sun. Blue light signals to plants that they are in the sun and should open their pores, take up carbon dioxide and start photosynthesizing.

“Insects are very sensitive to blue light and even more sensitive to ultraviolet light,” Longcore said. One potential solution would be to light streets with LEDs in the yellow or green spectrum, although Warrant cautioned that these lights might still draw insects. More research is needed to determine the effect that certain wavelengths of lights may have on insects and plants.

There’s evidence to suggest that dimming lights in the late hours and then switching them off after a certain time may help, Davies said. Lights activated by motion sensors are also a better option than constant illumination.

Pollinators and pollinator decline have gotten a lot of attention lately, but mostly in terms of how farming practices disrupt the environment during the day. “If we don’t look at nocturnal ecological communities,” Longcore said, “we’re only getting half of the environmental story.”

Read More: