ESA REFORM / CASE STUDY: GRIZZLY BEARS & TRAILS
GUILTY BY FORMULA
How the Endangered Species Act closes trails without proving a single animal was harmed.
The law makes it illegal to harm an animal protected by the Endangered Species Act. You would expect that to require evidence about an animal or the harm. On national forest land it doesn't. The government runs a calculation. If the miles of road inside a grid square exceed a number, harm is legally established and access closes. No animal has to be found, injured, or even looked for.
The arbitrary number is called a surrogate, and surrogates run public land management across the West. What follows traces one of them from the study that created it to the roads it governs today. The species responsible for the most closed access from abusive application of surrogate data is the grizzly bear. The problem has nothing to do with grizzly bears.
6 Bears
The entire field sample behind the road-density limit now enforced across the Northern Rockies. Two biologists collared six female grizzlies in the late 1980s. The standard is the average of those six animals' home ranges. It is still in force in 2026.
0 Studies
Research showing that closing a road ever improved grizzly survival, population density, or mortality. In twenty-nine years, no one has tested whether the rule works.
0 Traffic Counts
Vehicle counts on the low-standard forest roads being closed. The Forest Service does not measure use before removing access, and admits the gap in its own travel analyses.
THE MECHANISM
How a Number Becomes Law
Nobody in Congress voted for a road-density limit. No statute contains one. The number arrived through a sequence that repeats across species and agencies, and the sequence is important for everyone to understand. Here it is, traced through the grizzly example.
01. A small study finds a pattern.
In 1997, two federal and state biologists published a report on grizzly bears and roads in the Selkirk and Cabinet-Yaak recovery zones. Their tiny sample included six collared female grizzlies. Four in the South Selkirks, two in the Yaak. Because the female data were thin, they added two males. The telemetry ran from 1989 to 1994 against road maps from 1990 and 1991.
They measured where the bears spent time. They did not measure whether the bears survived, reproduced, or died. And they said so, in writing.
"Statistics regarding use and availability can provide a basis for development of standards, but survival and reproductive success must be considered when selecting animals or results which may be the basis for standards."
The authors flagged their own small sample as a limitation on interpretation. They were describing where six bears walked, not what kept bears alive.
02. The agency converts the weak pattern into a standard.
IThe percentages now enforced across the Northern Rockies are the average composition of those six home ranges. The Fish and Wildlife Service states the lineage plainly in its Biological Opinion for the Kootenai Revised Forest Plan: the road density standards are based on the research of Wakkinen and Kasworm.
03. The number becomes a hard line in the forest plan.
Across the Northern Continental Divide Ecosystem, forests manage most bear management subunits with the objective of keeping less than 19 percent of a subunit above one mile of open road per square mile, less than 19 percent above two miles of total road per square mile, and at least 68 percent as roadless core. The Selkirk and Cabinet-Yaak forests run parallel standards set unit by unit.
04. The number replaces evidence of harm.
This is the step that turns a biology paper into a closure. Federal regulation permits an agency to express incidental take as a surrogate rather than as a count of animals. In grizzly consultations, that surrogate is road density. Exceed the number and take has legally occurred. Nobody has to produce a bear or an actual injury. It is all based off of weak and dated scientific assumptions.
05. Litigation enforces the number.
Organizations sue over plan consistency and biological opinions. Courts review whether the agency followed its own arithmetic correctly. They are not asked whether the arithmetic describes reality, because that is not the question in front of them.
06. Nobody rechecks.
In twenty-nine years.
A surrogate is supposed to stand in for harm that is real but hard to count. This one stands in for harm that has never been demonstrated at all. And grizzly deaths aren't hard to count.
Sources
Interagency Grizzly Bear Committee. (2021). Northern Continental Divide Ecosystem Grizzly Bear Conservation Strategy, at 41-43.
U.S. Department of Agriculture, Forest Service. (2011). Forest Plan Amendments for Motorized Access Management Within the Selkirk and Cabinet-Yaak Grizzly Bear Recovery Zones, Record of Decision.
U.S. Fish and Wildlife Service. Biological Opinion for the Kootenai Revised Forest Plan, Chapter II: Grizzly Bear, at 6-8.
Wakkinen, W. L., & Kasworm, W. F. (1997). Grizzly bear and road density relationships in the Selkirk and Cabinet-Yaak recovery zones. Interagency Grizzly Bear Committee, at 6-8, 17.
Interagency Cooperation Regulations, 50 C.F.R. § 402.14(i)(1)(i).
The Go-To Claim
"But Roads Bring People, and People Kill Bears"
This is the most frequent argument, and it falls apart pretty quickly.
Start with what is true. Human causes dominate grizzly mortality. In 2024, 70 of 72 documented deaths in the Greater Yellowstone Ecosystem were human-caused. Bear deaths do cluster near roads. Roads are how people reach places. None of that is in dispute.
The inference is where it breaks. "Access brings people, people kill bears, therefore close this road to these users" needs four things to be true that are not.
01. Mortality runs through conduct, not proximity.
Every documented pathway requires a specific act by a specific kind of user. Sort the record and it becomes obvious.
"More humans" is not a mortality mechanism. Specific conduct is. Every item on that list has a tool aimed directly at it, and not one of those tools is a mileage cap.
02. The logic gets applied to one user and no other.
If human presence is the harm, every access point is implicated. Paved highways carry the documented vehicle mortality and stay open. Grazing allotments generate the largest category of removals and stay open. Hunting seasons produce the misidentification and self-defense killings and stay open. Trailheads deliver people on foot deep into the backcountry and stay open, sometimes as the stated mitigation for closing the road.
Only the dirt road closes. The one category with no documented deaths behind it.
03. The standard counts miles. It does not count use.
A grown-in two-track counts the same as a graveled main haul road. One mile is one mile. Nothing in the formula asks how many vehicles use it.
The researchers most often cited on road effects reached the opposite conclusion. Northrup and colleagues, publishing in the Journal of Applied Ecology in 2012, measured grizzly behavior against actual traffic counts and recommended that roads with higher traffic volume should contribute more toward density thresholds than roads receiving less traffic.
The Forest Service cannot do that weighting, because it does not have the numbers. Its own national assessment of the forest road system states that precise traffic counting has not been accomplished on any forest for many years. The Helena National Forest roads analysis inventories 958 miles of maintenance level 2 road and lists estimates of use as information still needing to be collected.
The simplest fair demand in this fight
Count the traffic before you close the road.
The agency is removing public access on the theory that use of a route harms a species, without ever measuring the use.
04. A gate does not remove the human. It removes the law-abiding human.
The closure side proved this themselves. In two 2025 rulings, the Ninth Circuit and the District of Montana found the Forest Service had failed to account for chronic unauthorized use of closed roads, and for breached gates and barriers that were not promptly repaired. The remedy those plaintiffs seek is more decommissioning.
Read the finding again. Closures are being defeated by people willing to break the law, and the proposed answer is to permanently remove the roads that law-abiding users depend on. Unauthorized use is an enforcement problem with an enforcement solution.
The Sophisticated Version of This Argument
"Bears are drawn to quiet roads, and that is exactly where they die."
Some researchers argue that low-traffic roads function as an ecological trap. Bears are attracted to good habitat along quiet roads, and human access then kills them there, so a bear's willingness to use those roads is evidence of danger rather than safety.
That argument comes from a 2017 study in southeast British Columbia, and the trap it documents is specific: a valley bottom where concentrated fruit resources sat alongside concentrated human residence. The same research team's 2023 follow-up itemized what actually killed their collared bears and prescribed the fix. Secure attractants on private property. Reduce collisions with trains and vehicles. Not forest road density.
A trap requires an attractant and a lethal agent in the same place. A gated road twenty miles from pavement, with no residences, no garbage, no orchards, and no rail line, has neither. The trap has to be demonstrated in that setting. It has not been.
Sources
Center for Biological Diversity v. U.S. Forest Service (9th Cir. Feb. 24, 2025) (Black Ram Project, Kootenai National Forest).
Center for Biological Diversity v. U.S. Forest Service, No. CV 22-91-M-DLC, Order on Summary Judgment (D. Mont. Oct. 27, 2025) (Knotty Pine Project).
Interagency Grizzly Bear Study Team. (2025). Yellowstone Grizzly Bear Investigations 2024. U.S. Geological Survey, Tables 14-15.
Lamb, C. T., et al. (2017). Forbidden fruit: Human settlement and abundant fruit create an ecological trap for an apex omnivore. Journal of Animal Ecology, 86(1), 55-65.
Lamb, C. T., et al. (2023). Grizzly bear mortality and mitigation in a multi-use landscape. Conservation Science and Practice.
Northrup, J. M., et al. (2012). Vehicle traffic shapes grizzly bear behaviour on a multiple-use landscape. Journal of Applied Ecology, 49(5), 1159-1167.
U.S. Department of Agriculture, Forest Service. (1998). National Forest Road System and Use.
U.S. Department of Agriculture, Forest Service. (2004). Helena National Forest Roads Analysis Report.
What the Record Shows
The Cause-of-Death Ledger Has No Entry for Recreation
The Interagency Grizzly Bear Study Team assigns a cause to every known grizzly death in the Yellowstone ecosystem. Livestock removals. Conflict and garbage removals. Self-defense shootings. Misidentification during hunting season. Highway and rail strikes. Poaching.
Recreational use of a forest road is not a small entry. It is not an entry.
The historical record has the same shape. Across 74 human-caused Yellowstone grizzly deaths between 1992 and 2000, defense of life and property accounted for 43 percent and removal of conflict bears for 28 percent.
In the Cabinet-Yaak, the recovery area with the most detailed road standards in the system, the human-caused mortalities recorded from 2019 through 2024 break down as self-defense, management removals, poaching, cases still under investigation, and a single vehicle collision.
Earthjustice, the firm that litigates these closures, names the leading direct causes as livestock conflicts, encounters with bear hunters, and vehicle collisions. Recreational access is absent from their list too.
The Question No Agency Has Answered
Name the mortality pathway.
Before a route closes in the name of a species, the agency should be able to state how lawful recreation on that route kills or injures the animal. Not how it might. How it does. Twenty-nine years into the standard, the record contains no answer.
Sources
Gunther, K. A., Haroldson, M. A., Frey, K., Cain, S. L., Copeland, J., & Schwartz, C. C. (2004). Grizzly bear-human conflicts in the Greater Yellowstone ecosystem, 1992-2000. Ursus, 15(1), 10-22.
Interagency Grizzly Bear Committee. (2025). Cabinet-Yaak Grizzly Bear Recovery Area 2024 Research and Monitoring Progress Report, Table 1.
Interagency Grizzly Bear Study Team. (2025). Yellowstone Grizzly Bear Investigations 2024. U.S. Geological Survey, Tables 14-15.
Nobody Ever Checked
The Rule Has Never Been Tested. Its Own Advocates Admit It.
Set aside whether the standard is fair. Ask a simpler question. Does it work?
There is no before-and-after study. No controlled test. Nobody has closed roads in one area, left them open in a comparable area, and measured whether the bears did better. Why use the scientific method when you can get the policy outcome you want with guesswork? Twenty-nine years of closures have produced no evidence that closures produce bears.
That is not a BlueRibbon Coalition characterization. It is the opening admission of the paper most often cited to defend road closures.
"Management agencies frequently attempt to reduce human-caused mortality by managing road density and thus human access, but the effectiveness of these actions is rarely assessed."
Rarely assessed. Written by the researchers whose work gets cited to justify the closures.
And the threshold itself will not sit still.
The strongest modern paper behind the sub-one-mile line produced two different answers from the same data. If the model assumes females with cubs suffer reduced survival near roads, the threshold is 0.75 kilometers per square kilometer. If it assumes survival is similar regardless, the threshold is 1.25. One analytical choice, made by the analyst, moves the line by 67 percent.
0.75
km/km² threshold under the restrictive assumption. This is the version that supports the rule.
1.25
km/km² threshold under the alternative assumption, from the same study and the same data.
+67%
Distance between them, decided by one input the data cannot resolve.
What That Means
Two engineers, one road, two speed limits.
Two engineers study the same crash data on the same road. One assumes a particular kind of driver faces higher risk and sets the limit at 45. The other declines that assumption and sets it at 75. Same road, same data, same crashes.
The agency adopted the equivalent of 45 and wrote it into forest plans across four states. Nothing in the data settles which engineer was right.
And Yellowstone recovered at the road network the formula says should have caused population decline.
The Greater Yellowstone grizzly population was listed in 1975 at a few hundred animals. The 2024 estimate is approximately 1,050 bears, the largest since monitoring began in the 1950s. The federal recovery criterion is a minimum of 500. Occupied range grew from roughly 23,000 square kilometers in 1990 to more than 70,000 by 2020.
Binding numeric road standards did not reach the Yellowstone forests until April 2006, by which point the population had already passed 600.
Greater Yellowstone Ecosystem · 1975–2024
A record population, reached at a road network the standard was written afterward to freeze.
1,050
Estimated grizzly bears in 2024. The largest population since monitoring began in the 1950s.
500
The federal recovery criterion. Exceeded for more than two decades.
2006
Year binding numeric road standards reached these forests. The population had already passed 600.
The Forest Service explained its own baseline choice in its environmental impact statement: "The year 1998 was chosen as the baseline because this was the access level at which the grizzly bear population recovered."
That is the agency conceding the 1998 road network was compatible with recovery. The recovery came first. The standard was written afterward to freeze the roads at the level that existed while it happened.
Sources
Boulanger, J., & Stenhouse, G. B. (2014). The impact of roads on the demography of grizzly bears in Alberta. PLOS ONE, 9(12), e115535.
Interagency Grizzly Bear Study Team. (2025). Yellowstone Grizzly Bear Investigations 2024. U.S. Geological Survey.
Lamb, C. T., Mowat, G., Reid, A., Smit, L., Proctor, M., McLellan, B. N., Nielsen, S. E., & Boutin, S. (2018). The effect of habitat quality and access management on the density of a recovering grizzly bear population. Journal of Applied Ecology, 55(3), 1406-1417.
U.S. Department of Agriculture, Forest Service. (2006). Record of Decision, Forest Plan Amendment for Grizzly Bear Habitat Conservation, Greater Yellowstone Area National Forests.
U.S. Department of Agriculture, Forest Service. (2006). Final Environmental Impact Statement, Grizzly Bear Habitat Conservation, Issue 10: Grizzly Bear.
The Ratchet
Access Only Moves One Direction
The formula does not sit still on the landscape. Every mechanism in the system makes open mileage easy to lose and nearly impossible to regain.
Closures happen, project by project.
On the Kootenai National Forest, the Grizzly Vegetation and Transportation Management decision authorized 15.4 miles of active road decommissioning, 15.5 miles placed into stored service to improve habitat, and 27 miles of passive decommissioning. On the Idaho Panhandle, forest monitoring documents 26 miles decommissioned in 2010 and 38 miles in 2011, with named roads inside the Kalispell-Granite and Lakeshore bear management units, tied to Access Amendment compliance timelines.
15.4 mi
Active decommissioning authorized in a single Kootenai decision, plus 15.5 miles to stored service and 27 miles passive.
64 mi
Decommissioned on the Idaho Panhandle across 2010 and 2011, including named roads inside grizzly bear management units.
0
Published running total of miles closed in the Northern Rockies under grizzly road standards. The agency does not report it.
There is no published cumulative figure for the Northern Rockies. The agency closes roads under this standard without reporting the total, which is itself worth noticing.
Going the other way requires a federal case.
In March 2025, a federal court in Idaho enjoined a project on the Idaho Panhandle because the Forest Service had raised a road-mileage baseline through an administrative change instead of a formal plan amendment. The baseline in that unit is 316.4 total miles and 314.4 open. The agency wanted 340.0 and 337.4. The court held that increase requires a full forest plan amendment, with the environmental analysis and consultation that entails.
And a gate is no longer enough.
In October 2025, a federal court in Montana held that the agencies had violated three separate statutes by leaving unauthorized and user-created road use out of the density calculation, and found that any closure failing to effectively prevent motorized access also fails to comply with the standard. The Ninth Circuit reached the same doctrine in February 2025, making it binding across the region.
The practical effect is a push from gating toward physical removal. A gated road that gets breached now counts against the forest's density budget, so the compliant answer is to pull the culverts, rip the surface, and take the road off the ground permanently.
Sources
Alliance for the Wild Rockies v. U.S. Forest Service (D. Idaho Mar. 31, 2025) (Hanna Flats Project, Idaho Panhandle National Forests).
Center for Biological Diversity v. U.S. Forest Service (9th Cir. Feb. 24, 2025) (Black Ram Project).
Center for Biological Diversity v. U.S. Forest Service, No. CV 22-91-M-DLC, Order on Summary Judgment (D. Mont. Oct. 27, 2025) (Knotty Pine Project).
U.S. Department of Agriculture, Forest Service. (2010). Grizzly Vegetation and Transportation Management Project, Record of Decision, Kootenai National Forest. 75 Fed. Reg. 63,527 (Oct. 15, 2010).
U.S. Department of Agriculture, Forest Service. (2012). Idaho Panhandle National Forests FY2010-2011 Forest Plan Monitoring and Evaluation Report.
A gate can be reopened when the science improves. A decommissioned road is gone.
Delisting & Reform
Delisting Is Step One, Not the Finish Line
Delisting is currently on the table and should be pursued. The Yellowstone and Northern Continental Divide populations have met their recovery criteria, and a species that has recovered should come off the list. That fight is worth having and BRC is in it.
It is also not sufficient, and members deserve to know why before they celebrate.
Road-density standards were not imposed by the Endangered Species Act. The ESA specifies no road-density number anywhere. Every standard in force was adopted by the Forest Service as a forest plan amendment under a different statute, the National Forest Management Act. The Selkirk and Cabinet-Yaak Access Amendment in November 2011. The Yellowstone area habitat amendment in April 2006, whose Record of Decision cites the forest planning regulations directly.
Delisting is a Fish and Wildlife Service action under the ESA. It does not amend a Forest Service land management plan.
What Delisting Ends
The engine that generates closure litigation
Section 7 consultation. The biological opinions. The incidental take statements. The take surrogate itself. The legal hook that has driven two decades of road cases disappears.
What Delisting Leaves Standing
Every number already in the forest plan
The road-density standards remain binding on every project, until the Forest Service amends each plan through its own multi-year process. The gates do not open.
That is why the ask has to run on two tracks. Delisting removes the engine that generates new closure litigation, which is real and worth winning. Forest plan revision is what actually reopens ground. Working one without the other leaves the job half done.
Sources
Greater Yellowstone Coalition v. Servheen, 665 F.3d 1015 (9th Cir. 2011).
National Forest Management Act, 16 U.S.C. § 1604(i).
U.S. Department of Agriculture, Forest Service. (2006). Record of Decision, Forest Plan Amendment for Grizzly Bear Habitat Conservation, Greater Yellowstone Area National Forests.
U.S. Department of Agriculture, Forest Service. (2011). Forest Plan Amendments for Motorized Access Management Within the Selkirk and Cabinet-Yaak Grizzly Bear Recovery Zones, Record of Decision.
The Pattern
Different Species, Different Desert, Same Machine
This is not a grizzly problem. It is how the statute is being administered, and it will keep producing the same result until the surrogate itself is fixed.
West Mojave · Desert Tortoise
Eight tortoises a year, on professional judgment
FWS estimated take from casual off-highway use across 3.1 million acres at roughly eight adult tortoises per year, about 0.07 percent of the population, and called the estimate professional judgment rather than field study. Eleven years of monitoring found five. A 2024 court order produced closures affecting roughly 2,200 miles of route.
Northern Rockies · Grizzly Bear
Six bears, and nobody checked
Six collared grizzlies produced the threshold in 1997. No documented recreational mortality appears anywhere in the federal record. No study has ever tested whether closing a road helps the species.
Two ecosystems, two species, two unrelated bodies of science, and the same structure underneath. An arbitrary number stands in for proof. The closure lands on recreation every time. And in both cases, the things actually killing the animal go unaddressed.
Sources
Center for Biological Diversity v. Culver, Case No. 3:21-cv-07171-SI, Order on Cross-Motions for Summary Judgment (N.D. Cal. Oct. 22, 2024).
U.S. Fish and Wildlife Service. (2019). Biological Opinion for the West Mojave Route Network Project. FWS Administrative Record 5405-5406.
Wakkinen, W. L., & Kasworm, W. F. (1997). Grizzly bear and road density relationships in the Selkirk and Cabinet-Yaak recovery zones. Interagency Grizzly Bear Committee.
Bottom Line
Closing roads on a standard drawn from six bears, never tested in twenty-nine years, against a mortality cause that appears nowhere in the record, is not conservation.
The grizzly deserves science aimed at what is actually killing it. Recreationists deserve an honest accounting of why their access is being taken. Right now, neither is getting one.
What This Means for OHV Access
The Species Is the Instrument. The Closure Is the Result.
BlueRibbon Coalition has never argued that grizzly bears do not deserve protection. They do. But recovery has to be built on evidence, and the regulatory response has to match the conduct that actually kills the animal.
Closing a dirt road is easy to draw on a map and easy to defend in court, and happens to perfectly align with the policy agenda of anti-access groups. Livestock conflict prevention, attractant management, hunter identification standards, anti-poaching enforcement, and highway crossing structures are slower, cost money, and require confronting interests with more political weight than recreational riders. The easy option got chosen. That is political opportunism, not conservation.
The Origin
Six bears set the line
The threshold traces to a 1997 report on six collared female grizzlies whose authors warned in writing that survival and reproduction had to be considered before their numbers became standards. They were not.
The Test
Twenty-nine years, no study
No before-and-after research has ever shown that closing a road improved grizzly survival, density, or mortality. The leading paper cited for closures concedes effectiveness is rarely assessed.
The Evidence
No entry in the ledger
The federal cause-of-death record contains livestock removals, self-defense shootings, misidentification, highway strikes, and poaching. Lawful recreational road use does not appear.
The Proof
Yellowstone recovered anyway
The population reached a record high with the road network in place, and the Forest Service chose 1998 as its baseline because that was the access level at which the population recovered.
A surrogate that has never been validated is not best available science. It is an assumption driven by an agenda.
Take Action
Fix the Formula
The Fish and Wildlife Service published a proposed revision to the grizzly bear listing and its Section 4(d) protective regulations on July 17, 2026. Comments close August 17, 2026, under docket FWS-R6-ES-2024-0186.
Because grizzlies are listed as threatened rather than endangered, the take prohibition reaches them only through what the Service writes into this rule. That is real discretion. The current draft says nothing about roads, recreation, or travel management, and that silence is where the formula survives.
One point of precision, to also consider. In July the Service and NOAA Fisheries finalized a rule narrowing what counts as take under Section 9, effective September 14, 2026. Road-density surrogates operate under Section 7, which that rule does not touch. Reform that stops at Section 9 leaves the road formula fully intact. That is exactly why this rule has to reach it.
What BRC is asking the Forest Service to do:
01.
State explicitly that lawful motorized recreation on designated routes, routine road maintenance and decommissioning, and federal travel management decisions do not constitute take of a grizzly bear.
02.
Stop keying take to linear road-density metrics that have never been validated. Where an agency uses a surrogate, require that it be weighted by actual measured use, as the underlying research itself recommends.
03.
Require an identified mortality pathway before recreational access is restricted. If an agency cannot state how lawful recreation on a specific route kills or injures a specific animal, that route should not close in the name of the species.
04.
Finalize the single lower-48 distinct population segment, which gives the Service a legally durable path to revise the listing where recovery has occurred.
04.
Acknowledge that forest plan road-density standards were adopted under the National Forest Management Act, that they survive delisting, and that they must be re-examined on their own merits rather than left in place as permanent residue of a listing decision.
TAKE ACTION
Real Conservation Demands Real Science
Off-roaders deserve an honest accounting.
It's Time For a New Playbook.
BlueRibbon Coalition isn't standing idly by. We're fighting to change the rules that make cooperating in the current system a losing endeavor. Join us.