The number everyone quotes for American roadside habitat comes in two sizes. A 2019 paper in Frontiers in Ecology and Evolution describes state departments of transportation as managing “an estimated 17 million acres of potential habitat for monarchs,” attributing the figure to a 2014 assessment by Ament and colleagues. A 2021 literature review prepared for the Maine Department of Transportation puts it at “approximately 10 million acres of adjoining roadside land,” attributing that to Forman and colleagues in 2002 and to Wojcik and Buchmann in 2012. The same sentence in the Maine review gives the road network as 17 million miles, which is a good reason to read both figures as a citation trail rather than as a measurement of anything.
Nothing in the ecology depends on resolving that gap. The policy does, because every argument for managing roadsides as habitat rests on the area being large enough to matter, and both figures reach the reader through a chain of citation rather than from a survey. That shape is familiar here. As wildlife crossings shows, the trustworthy numbers in roadside ecology are corridor-specific, and the national picture is assembled from estimates.
Two grant programs, pointed in opposite directions
The Infrastructure Investment and Jobs Act created a dedicated roadside pollinator program, now 23 U.S.C. 332, directing the Secretary to fund activities that benefit pollinators on roadsides and highway rights-of-way, “including the planting and seeding of native, locally-appropriate grasses and wildflowers, including milkweed.” Eligible applicants are state transportation departments, Indian tribes and federal land management agencies. The federal share is 100 percent, no individual grant may exceed $150,000, and the authorization is $2,000,000 for each of fiscal years 2022 through 2026. Caltrans, summarizing the competition for California applicants in May 2024, recorded $3 million available in that round, a $150,000 ceiling and no cost sharing required.
The same law created a second roadside vegetation program in section 11522, this one to eliminate or control invasive plants along and adjacent to transportation corridor rights-of-way. Its authorization is $50,000,000 for each of the same five fiscal years. The federal share is 75 percent where a project revegetates with native plants and wildflowers, including pollinator-friendly ones, and 50 percent for anything else. Grant funds may not pay for mowing unless mowing is identified as the best means of treatment according to best management practices or is used in conjunction with another treatment.
Set side by side, the two authorizations show federal money for roadside vegetation running twenty-five to one in favor of removing plants over establishing them. Authorizations are not obligations, and the removal program’s own preference for native revegetation softens the contrast. The ratio still records which problem Congress treated as urgent.
The statute names the mowing practices, and the evidence is messier
Section 11528 is unusually specific for a funding statute. To qualify, an applicant’s pollinator-friendly practices plan has to describe practices such as reducing the mowing swath outside of the state-designated safety zone, increasing the mowing height, reducing the mowing frequency, refraining from mowing monarch and other pollinator habitat during periods in which monarchs or other pollinators are present, using a flushing bar and cutting at reduced speeds to reduce pollinator deaths from mowing, and reducing raking along the right of way.
The Maine review, which screened peer-reviewed and gray literature for practices applicable to forested northeastern landscapes, supports the frequency and timing items on that list and says little about swath width or cutting height. Its three consistent recommendations are to reduce mowing frequency and time mowing to pollinator activity, to target herbicide applications at undesirable species using backpack sprayers, and to plant native seeds, seedlings or shrubs while leaving some exposed soil for nesting bees.
Underneath that summary the studies disagree in ways worth knowing before writing a specification. Mid-season mowing benefits plant diversity while reducing butterfly abundance and species richness, an effect that partial mowing can mitigate. Late-season mowing does not reduce butterfly abundance, though the review notes that mowing at any point in the butterfly life cycle is likely to have some negative effect by removing plants used for feeding or egg laying. On plants the split is sharper still: some studies find that mowing allows non-native species to establish, and others find that mowing promotes native species. Less frequent mowing can maintain safe visibility and hold soil in place by keeping plant density up, which matters wherever the disturbed ground described in earthwork and grading has to be stabilized. And unmanaged right of way can develop diverse plant communities and grow into high-quality pollinator habitat on its own.
The monarch record, as the record states it
The Fish and Wildlife Service proposed on 12 December 2024 to list the monarch butterfly as a threatened species with a section 4(d) rule and to designate critical habitat, at 89 FR 100662. The extinction probabilities in that document are the figures most often flattened in retelling, and they are ranges attached to two separate populations. For the eastern migratory population the Service estimates a 24 to 46 percent probability of extinction in 30 years and 56 to 74 percent in 60 years. For the western migratory population it estimates 92 to 95 percent in 30 years, reaching 99 percent by year 60. The proposal’s own docket, FWS-R3-ES-2024-0137, is where the status of any final rule can be checked.
On causes, the proposal is more specific about milkweed than about anything a highway agency controls. Between 2008 and 2016 a total of 4.9 million acres of grassland were converted to new cropland, including up to 3 million acres of Conservation Reserve Program land. Past loss of agricultural milkweeds in the Midwest is associated with an estimated 81 percent decline in monarch production, in part because monarch egg densities were higher on milkweed growing in agricultural fields, at 3.89 times more eggs than on non-agricultural milkweed. That matters disproportionately to the eastern population because estimates of the share of overwintering monarchs originating in the Midwest crop belt range from 38 percent to over 85 percent.
Roadsides enter that document twice, and both passages are more careful than the advocacy around them. On mortality, the Service concludes that vehicle strikes are not one of the primary drivers of changes in monarch populations and that the impacts are considered minimal at a population or species level, while noting research suggesting there may be roadkill hotspots during migration. On habitat, it states that research suggests roadside monarch habitat “can still provide a net benefit to the species, despite losses due to collisions, through strategic improvements to roadside vegetation management.” The net-benefit finding is the load-bearing one for any transportation agency writing a plan, and it is a conclusion about strategic management rather than about roadsides in general.
Scale gives the same answer from the other direction. The Mid-America Monarch Conservation Strategy set a goal of adding 1.3 billion milkweed stems by 2038, which the Service describes as the target needed to support 14.8 acres of occupied overwintering forest for the eastern population. As of September 2024, state agencies had implemented milkweed restoration on more than 8,780,404 acres, adding an estimated 546 million stems nationwide. A roadside pollinator program authorized at $2 million a year nationally is a real contribution to a target measured in billions of stems while being far too small to close the gap on its own.
Nobody can score a roadside quickly, which is the practical bottleneck
The 2019 Frontiers paper began by surveying roadside managers, and the finding that shaped everything after it was that managers are often limited by time, funding and expertise in plant identification. The authors built a rapid assessment protocol and a habitat evaluator that turns field data into habitat quality scores, then field-tested it in Minnesota against the more intensive Integrated Monarch Monitoring Program.
Their milkweed densities show why single-season monitoring misleads. Using the intensive protocol, mean density for all milkweed species combined was 1,242 plants per hectare in 2017 and 2,807 in 2018, with a two-year mean of 2,052. Those are the same roadsides, a year apart, differing by more than a factor of two. A program judged on one year of data can be reported as a success or a failure depending on which year the surveyor showed up.
Where habitat management meets weed law
Iowa runs a long-standing version of this work through integrated roadside vegetation management, and the Tallgrass Prairie Center at the University of Northern Iowa describes what a county practicing it actually does: it sprays herbicides strategically to control the most problematic weeds, mows strategically for safety and weed control while leaving vegetation farther from the road unmowed, conducts prescribed burns to keep prairie vegetation healthy and stop trees and shrubs establishing, and protects remnants of unplowed prairie in the roadside. Funding runs through the state’s Living Roadway Trust Fund, which supports those activities for eligible cities, counties and applicants with statewide impact.
Much of that list is subtraction. The same tension is written into federal law. Section 11522 defines an invasive plant by naming cheatgrass, medusahead, Japanese honeysuckle, phragmites, autumn olive, Bradford pear, wild parsnip, sericea lespedeza, spotted knapweed, garlic mustard and palmer amaranth among others, and a crew funded to remove those plants works the same right of way, often in the same season, as a crew funded to establish flowering natives. The monarch proposal resolves the conflict in the same direction, listing among activities it would except from take prohibitions habitat restoration and management including mowing and haying, and “activities to eliminate plant communities that contain invasive plants or noxious weeds as part of site preparations or habitat enhancement activities.”
The certainty already purchased
In early 2020, before any of this funding existed, energy and transportation entities finalized the Nationwide Candidate Conservation Agreement for Monarch Butterfly on Energy and Transportation Lands. Enrolled participants create and maintain habitat along their rights of way and carry out conservation measures, and in exchange the Service assures them they will not have to implement additional measures should the species be listed.
That agreement predates the listing proposal by nearly five years, which makes the roadside the one monarch habitat whose managers already know what a listing would cost them. Grant programs expire on five-year authorizations and mowing calendars get rewritten whenever a state changes maintenance contractors. Of all the arrangements gathered in the environment section, an assurance signed against a species that is still only proposed for listing is the most durable, and it was secured by the sector whose direct contribution to the decline the Service rates as minimal.