The wind races through the grass as Marla Spivak swings her insect net across a patch of thistles in a roadside ditch.
She kneels to carefully examine the catch.
"Lots of flies, there's a grasshopper, there's a dragonfly," Spivak said.
But no bees. And that's not surprising. The roadside ditch has a cornfield on one side and a just-harvested wheat field on the other. There's not much for bees to eat.
"In some of the locations where there's a lot of flowers you see a lot of very cool native bees," Spivak said. "But this one we call bad because we kind of presume we're not going to find a lot of native bees."
The University of Minnesota entomology professor is working with a graduate student to catalogue native bees found on different types of landscapes.
Spivak's work is one of several related studies in eastern North Dakota this summer. Researchers are trying to better understand why so many bees are dying all across the country.
North Dakota is a good place to do it because more than half the bees in the country are found in the Dakotas and western Minnesota. North Dakota leads the nation in honey production, and there are about 2.5 million bee colonies in the country.
Many of those bees are moved to California each winter. It takes 1.5 million colonies just to pollinate the almond crop each year.
The bee population has been in decline for many years. There were twice as many honey bee colonies in the 1950's as there are today.
Many people believe pesticides are the reason bees are dying.
Spivak said the current research is the first in depth look at how landscapes might be affecting bee health. She said people have changed the landscape; here are fewer flowering plants, and bees might not be getting a healthy diet.
"They have to get all their amino acids from different flowers so they need a diversity of protein which is there pollen," she said. "And if their nutrition is bad then they are much more susceptible to the effects of pesticides, ... diseases and parasites."
Another graduate student working with Spivak will collect honey bees from hives in good and bad locations. Those bees will be carefully studied to see how different kinds of pollen affect the bees' immune systems.
A few miles away, researchers decked out in white clothing and head nets are opening hives to check honey bee health.
In the largest bee study in the country, 288 hives will be followed for three years, said Jeff Pettis, research leader for the Agriculture Department's bee research lab in Beltsville Maryland.
Pettis said with all of the research into why bees are dying, the importance of bee diet has been somewhat overlooked.
"Honey production is only one side," he said. "They need pollen to raise young bees. So even if they had a great nectar source to produce honey if there's not a diversity of pollen out there the bees won't do well."
Pettis said some people want to blame bee deaths on pesticides. Others focus on parasites like the varroa mite. He said the real reason is likely much more complex and might start with nutrition. Bees that eat only one or two kinds of pollen, like a person eating mostly junk food, will have weakened immune systems.
"And that sets up what I call a cascade effect. Now you've got a little bit of varroa stress, get exposed to a little bit of pesticide," Pettis said. "Well what was the leading cause? They were all a factor in bringing the colony down. That's part of the reason I'm kind of excited about this study."
The research project isn't just about bees. The U.S. Geological Survey is also involved in detailed mapping of the landscape around bee hives. As bees stay within two miles of the hive, researcher Ned Eulis is carefully mapping all the plants in that bee zone. The dash of his truck is covered with hand drawn maps.
Eulis said everyone agrees bees need a diverse habitat. But no one really knows exactly what mix of plants bees need to be healthy.
If you look at the road ditches, there's milkweed here, there's going to be alfalfa, some native plants as well," he said. "How much is enough? We don't know the answer so we're really at a very fundamental stage."
Researchers hope what's learned here on the North Dakota prairie will lead to changes in farm policy, encouraging farmers and others to create landscapes that are more bee friendly.
They expect to have some answers later this year, and in three years they will have perhaps the most detailed understanding of bee nutrition ever assembled.