The two fields in southern Minnesota looked abandoned at first glance.
Rows of solar panels stood over what had been cropland, and between the rows, seed had just gone into bare, worked over soil.
Bare ground is a slow start for a prairie.
But the counts under those panels kept climbing, and the steepest climb of all belonged to an insect nobody manages.
What a solar field gives a bee that a cornfield cannot
A working cornfield offers a wild bee almost nothing. The canopy closes fast, the ground between stalks stays bare, and the weed control program makes sure little else flowers. A solar array on retired farmland is a different proposition entirely, because the ground under and between the rows is never harvested and can be planted to grasses and flowering forbs instead.
The mechanism is a simple one. Bees need flowers across the whole season and undisturbed ground or stems to nest in, and a seeded solar field supplies both where row crops supplied neither. Researchers have also proposed that partial shade under tilted modules shifts bloom timing and stretches the flowering window, but that idea was not what this work set out to test.
One correction worth making up front: most native bees are not hive insects. The great majority are solitary, nesting in soil or hollow stems, so a rise in native bee numbers means many separate nests succeeding rather than one colony swelling.
What showed up between the rows
The early seasons were modest. Prairie takes time to find its feet, the flowers were sparse, and the insects were scarce along with them. It would have been easy to conclude that the whole idea was a fine thought nature had declined to confirm.
Then the numbers moved. Total insect abundance tripled, flowering plant species richness rose about sevenfold, and insect group diversity increased by an average of roughly 13 percent a year. Native bee abundance climbed more than twentyfold, the steepest change anywhere in the dataset.
Beetles, hoverflies and moths were the most numerous groups counted. The second result was the one that left the fence line behind: bees were also working soybean flowers in the crop fields next door, which is where the finding stops being a story about a solar site and starts being a story about the farm beside it.
How the counting was done near the prairie edge
The two sites, both built on retired agricultural land, were planted with native grasses and wildflowers in early 2018. From August of that year through August 2022, survey teams walked transects repeatedly each summer, logging 358 transect observations in the habitat areas.
Those surveys detected 10,943 beneficial insects spanning four orders: bees and wasps, flies, moths and butterflies, and beetles. Goldenrod soldier beetles alone made up about 35 percent of observations, followed by syrphid flies and moths.
The authors describe what they saw as “relatively rapid” insect community responses to grassland restoration, which set up the more surprising result waiting in the neighboring fields.
Why the soybean result is the odd part
Soybeans are largely self pollinating, so most growers do not treat bee visits as a yield input. The study did not measure soybean yield. What it measured was visitation, and bee visits to soybean flowers beside the solar habitat were comparable to visits beside grassland enrolled in the Conservation Reserve Program, and greater than visits inside the soybean field or along the roadside.
That comparison is the load bearing part, because it puts a working power plant roughly on par with land paid to sit in conservation cover. The authors argue that siting arrays on marginal farmland and seeding them could preserve prime farmland while supporting pollination and pest control nearby.
The scale question is not small: the Department of Energy estimates more than 10 million acres will be needed for solar by 2050, and the American Farmland Trust projects that over 80 percent could be sited on agricultural land. A similar dual use question comes up where cranberry bogs share the same ground with solar panels.
What holds it back and what still needs measuring
This result does not transfer automatically to every solar site. Panel height matters, because a rack set too low shades tall wildflowers out before they establish. Soil preparation matters just as much, since compacted, long farmed ground resists native seed for years. Parts of these sites were research plots that received additional seeded species and different management from the rest of the array.
The design has limits too. It followed two sites over time without an unplanted control, so it shows that conditions improved rather than isolating exactly why, and both began from the very low ecological baseline of retired cropland. A developer placing arrays on intact prairie would start from a higher point and could easily move the other way.
Active management each season is the difference between a living meadow and a gravel lot with panels above it. Even so, the direction holds. The five year study in Environmental Research Letters reports gains in every habitat and biodiversity metric tracked, plus higher bee visitation in neighboring soybeans. A related question runs through reporting on wind turbine towers and the insects drawn to them: what lives near these machines depends largely on what gets planted at their feet.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
