The midges of Lake Myvatn — literal translation, 'Midge Lake' — in northern Iceland make up two-thirds of the lake's biomass, and often form swarms that hover like clouds over the surface of the water. Anthony Ives, a theoretical ecologist at the University of Wisconsin–Madison, puts their abundance into perspective by describing it as being “like going into prairie grasslands and having the majority of the animals there be a single species of grasshopper”. But despite their dominance, the numbers of these tiny insects fluctuate wildly. And a mathematical model Ives developed to explain the population dynamics of the lake's midges suggests that they are extremely sensitive to both natural and human-induced change.

Ives began collaborating with Árni Einarsson and Arnthor Gardarsson at the University of Iceland in Reykjavik 10 years ago. Since 1977, these researchers had been gathering data on the population density of the midges, Tanytarsus gracilentus, living in Lake Myvatn. The data showed that the abundance of these midges, which feed on algae, fluctuates by almost six orders of magnitude. “The thing that really struck me was not just the extent of these fluctuations, but the fact that they're not random, nor are they regular. They're something in between,” says Ives. Although some other animals also show dramatic population outbreaks, these typically occur in a regular fashion; outbreaks and crashes of Lake Myvatn's midges occur irregularly, 4 to 7 years apart.

To understand the phenomenon, Ives cobbled together statistical tools normally used to monitor stock-market activity or the trajectories of interplanetary satellites, and applied them to Einarsson and Gardarsson's field data. The model showed unusual mathematical properties, so they enlisted Vincent Jansen at Royal Holloway, University of London, Epsom, to help interpret what these meant. In the model they designed, which is described on page 84, population abundance shifts between a constant state and one that is cyclical.

“It's this shifting from one type of dynamic [constant] to another type [cyclical] that can produce high-amplitude cycles and make the frequency of these cycles unpredictable or irregular,” says Ives. Even small changes in the weather, or other environmental events, including those caused by human activities, may result in wild fluctuations in population density, he adds.

Ives's model also helps to explain the impact that historical dredging for the mineral diatomite may have had on the lake. The operation, which started in 1967, was abandoned in 2004 after becoming what Ives describes as “an environmental cause célèbre” when the fish populations started to diminish drastically. His mathematical model suggests that such dredging could have increased the size of the fluctuations in the midge population. Because these organisms are the main source of food for Lake Myvatn's fish, a crash in the midge population would have left fish with nothing to eat.

Ives hopes to continue this collaboration, and expand it to include other researchers, allowing them to gather further basic ecological information about the midges and the entire ecosystem that they affect.