Leave your car parked outside in Miami, the US, for too many days in August, and you're likely to see a fine film of orange powder coating the windshield.
It's dust - African dust - carried thousands of kilometres across the Atlantic to parking lots in Florida.
This phenomenon, known as the Saharan Air Layer, brings several hundred million tonnes of dust to the Americas each year. It helps build beaches in the Caribbean and carries nutrients that fertilise the Amazon rainforest. The dust is also thought to help suppress hurricanes along the coast.
It has been happening for thousands of years, but in the past few decades, scientists have become concerned about the effects of this dust on coral reef ecosystems in the Caribbean.
Studies have linked the dust plumes to toxic "red tides" - huge algae blooms that poison fish and trigger skin and breathing problems in people.
There's also some evidence suggesting that African microbes may be hitching rides on the dust and poisoning corals on the other side of the sea.
But what about the bacteria that already live in the ocean? In a study published this week in the Proceedings of the National Academy of Sciences, scientists examined how one genus of bacteria, a pathogen called Vibrio, responded to the influx of dust that rains down on the Caribbean in the late northern summer months.
The answer: With gusto.
"Bacteria are really the first rapid responders to what's coming in," said William Landing, professor of Earth, ocean and atmospheric science at Florida State University and a co-author of the study. Vibrio bacteria are able to snatch up nutrients blown in with the dust before other microorganisms.
On a heavy dust day, the bacteria immediately start capitalising on the sudden influx of iron, which is otherwise hard to come by in the ocean. Within 24 hours, they grow into vast blooms. Before the dust, Vibrio bacteria may represent 1 per cent of all microbes in the water, but that proportion swells to nearly 20 per cent afterward.
"A lot of work has gone into trying to understand what happens to a marine ecosystem under a dust event, but most of that work has been on phytoplankton and algae" - the culprits behind the red tides - said Erin Lipp, a microbiologist at the University of Georgia who was also part of the study. "But no one has really looked at how bacteria might be responding to the dust."
This study provides a model of how bacteria respond to the influx of dust, she said. And that will let Lipp and her colleagues start asking the big questions that naturally come next:
"What does it mean?" she said. "Does it help or hinder primary productivity? What does it mean when you have these blooms of vibrios that could potentially be pathogenic to humans and corals?"
Landing said it's too soon to say whether the bacteria blooms are good for the ecosystem. In fact, he resists the urge to qualify them as "good" or "bad."
"It's not 'bad' or 'good'; it's just a component of the nutrient cycle that they're sitting in the middle of," he said.
Although humans are not likely to want to go swimming or eat fish from water with Vibrio blooms (the bacteria can cause diarrhea, wound infections and other unpleasantness), its presence may be beneficial in some cases. Landing pointed out that the bacteria blooms may help feed phytoplankton, photosynthetic microbes that absorb carbon dioxide from the atmosphere.
On the other hand, Vibrio bacteria are toxic to the Caribbean's already-stressed corals, which we would very much like to protect. It can contribute to bleaching and destroy tissue, and although Vibrio bacteria are usually never present in large enough numbers to pose a major threat, under the right environmental conditions (such as those created by a dust-borne surge in iron) it will overwhelm reefs' defences.
Gene Shinn, a retired US Geological Survey scientist now based at the University of South Florida, said he was "delighted" to see Lipp and Landing's study.
Shinn, who was not involved in the study but has worked with some of its authors in the past, has been sounding the alarm on African dust in the Caribbean for more than a decade, ever since he noticed that the first major Caribbean coral die-off began in 1983 - the same year that a catastrophic drought struck North Africa, turning hectares of farmland there into dusty desert.
In a later study, Shinn showed that Aspergillus, a soil fungus that doesn't reproduce in sea water but is endemic to Africa, was killing off Caribbean sea fans.
The bacteria blooms, he said, are further evidence that African dust is something to concern marine scientists. Exactly how worried they should be is still unclear.
"Bacteria have always been there," Landing said. "It's only if we get them more frequently that it's going to cause a problem."
Much of that depends on how climate change affects the dust in the Saharan Air Layer, and the science on that is still unsettled. On one hand, desertification of the Sahel, the arid transitional region at the edge of the Sahara that provides much of the dust, could increase the amount of dust in the air. But other studies have suggested that climate change will cause that dust to linger over Africa, rather than crossing the Atlantic, because of changes to regional wind patterns.
"That's something we're still looking into," Lipp said. "This study was about homing in on what's happening."
Now, she added, the goal is to figure out what's going to happen next.