Landscape fabric is sold as a labor saving product.
Roll it out, cover it, and nothing has to be weeded again for years.
What it actually does is seal a surface.
Underneath it there is soil that no animal can enter, and above it a layer of decaying mulch that eventually grows the weeds the sheet was bought to stop.
Somebody finally tested what happens when you take the lid off instead.
So what is a bee actually looking for down there?
Everything depends on the top inch
Most bees are not social and do not build combs.
A female digs a tunnel, packs a cell at the bottom with pollen, lays an egg on it and seals the whole thing behind her.
Roughly 70 percent of native bee species in North America live that way, underground rather than in stems or cavities.
Her young then stay in that hole for as much as eleven months before they come out.
So the constraint is not flowers. It is an entrance she can physically make.
Compacted ground resists the dig, a sealed surface offers no opening at all, and a dense mat of roots is a wall she has no tool for.
Warmth matters too, because development underground runs on temperature, and open sunlit soil sits warmer and drier than the same ground under deep litter.
Five squares of ground, laid out five ways
The experiment ran across 21 sites in city and county parks and greenspaces around Columbus.
They were picked to span a gradient, from sites with barely 1 percent developed land within roughly 1,600 feet to sites at 88 percent.
Each plot measured about 10 by 16 feet and was cut into three identical strips.
Every strip carried five squares of about 11 square feet each.
One was tilled to bare ground. One got about three quarters of an inch of dead leaves and branches. One got drainage pebbles across half its surface. One got soil mounds about 10 inches high, compacted.
The fifth was the control, and it is worth being precise about it: not untouched ground, but vegetation trimmed to a standard low height.
What the traps and the watching turned up
Fieldwork ran across six visits from May through September of 2023, plus a return in April of 2024.
Each visit combined a 15 minute watch of the plot with emergence traps and pan traps left for 48 hours on days warm enough to fly, above about 70 degrees.
The haul was 170 nesting females, 26 species across eight genera, with nesting recorded at 18 of the 21 sites and four fifths of it from a single family.
Every manipulated square averaged more nesting than the control, and the trap data showed the bees were not merely favoring the disturbed squares but actively avoiding the vegetated ones.
The genera split by surface. Halictus turned up most in bare ground. Andrena took leaf litter and the mounds. Augochlorella went almost entirely to leaf litter.
Lasioglossum used everything except the control.
Colonization was strongest at the more urban sites, and at sites with the least bare ground nearby.
Four reasons to hold this loosely
It is a preprint, posted publicly and not yet through peer review.
One site produced 38 percent of all the nesting bees analyzed, which is the kind of concentration that can carry a result on its own.
The Andrena and Augochlorella samples were small enough that the surface preferences for those two are suggestive rather than settled.
And the gardening advice does not follow as directly as it looks.
Tilling created the winning surface here, but tilling the same patch every spring would destroy the brood already sitting under it, so the move is to open ground once and then leave it alone, the same logic that governs birds on farmland and the species that persist inside heavily managed landscapes.
Nobody has followed those nests through to emergence either, so the count is of females that started, not of young that finished.
The cheapest conservation is a subtraction
Most habitat advice asks you to buy or plant something.
This one asks you to stop doing three things: sealing soil under fabric, piling heavy wood chips deep enough to bury it, and mowing every square foot to the same height.
Conservation guidance is blunt about the fabric and the heavy chips, and it is the same guidance that warns against turning the soil, as the standing advice on bare ground makes plain.
The Columbus result adds the part gardeners get wrong.
A worn patch is not a failure of the lawn. It is the only door in the building, and ground nesting bees found the new ones inside a single season, as the posted study records.
The squares nobody scraped stayed almost empty.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
