Drill seeding usually restored more sagebrush—but location changed which method offered better value

The discovery

The researchers asked where aerially spreading sagebrush seed or placing it into the soil with a drill would provide more ecological improvement per dollar after wildfire across the western U.S. sagebrush biome.

The research question and why it matters

Long-running field records show that post-fire sagebrush recovery is slow and varies with climate, soils, invasive annual grasses and the treatment used. Earlier biome-wide work by members of this team modeled vegetation responses to restoration, while a separate 2025 analysis estimated how public-land treatment costs change with project size, slope and access. The new study links those two evidence layers, converting predicted ecological gains and costs into spatial estimates of improvement per dollar rather than assuming that the most effective or cheapest method is automatically the best investment.

What researchers found

Thirty years after the modeled fire, 28% of pixels under aerial seeding and 39% under drill seeding reached at least half of their estimated sagebrush recovery potential. Yet the modeled improvement over natural recovery was generally small: aerial seeding added less than one percentage point of cover in 92% of pixels and averaged a 0.34-point gain, while drill seeding added less than one point in 74% of pixels and averaged 1.51 points. Drill seeding produced greater sagebrush cover than aerial seeding in 86% of pixels, but it was always more expensive at a matched project size. For 5,000-acre projects, drill seeding had equal or greater cover-based cost-effectiveness across 70% of the modeled landscape and equal or greater recovery-probability cost-effectiveness across 92%. Aerial seeding offered better relative value in some places, particularly on steeper slopes.

Results at a glance

Key results from the tested systems

429,718 km²

modeled sagebrush landscape

The study area overlapped 12 western U.S. states and was evaluated in 300-meter pixels.

86%

pixels where drill seeding produced more cover

Greater modeled effectiveness came with consistently higher drill-seeding costs.

$13 vs. $44/acre

median cost for 5,000-acre scenarios

The modeled medians were $13 for aerial seeding and $44 for drill seeding, in 2021 dollars.

70% / 92%

landscape favoring drill by two value measures

For 5,000-acre scenarios, drill equaled or exceeded aerial seeding for cover-based cost-effectiveness across 70% of the landscape and recovery-probability value across 92%.

How the research worked

The team combined previously published spatial models of sagebrush cover and recovery 30 years after a hypothetical wildfire with models of per-acre treatment costs. The ecological models compared aerial seeding, drill seeding and natural recovery at 300-meter resolution. Cost models incorporated location, slope, road access, distance from urban areas, vegetation type and project size, expressed in 2021 dollars. For each pixel, the researchers divided the predicted treatment improvement over natural recovery by estimated cost per acre. They evaluated both change in sagebrush cover and change in the probability of reaching a recovery benchmark. One thousand random draws propagated uncertainty in cover and cost estimates, and generalized additive models tested how environmental and access conditions related to cost-effectiveness. Scenarios represented projects of 100, 1,000 and 5,000 acres.

Subjects or systemComputer model

Research designSpatial cost-effectiveness modeling study combining published ecological-recovery and treatment-cost models

Evidence baseA modeled 429,718-square-kilometer portion of the sagebrush biome across 12 western U.S. states, evaluated in 300-meter spatial pixels for hypothetical 100-, 1,000- and 5,000-acre aerial- and drill-seeding projects after wildfire

How to interpret this design

The result is conditional on the model structure, inputs, boundary conditions and scenarios chosen by the researchers. Agreement with known observations strengthens confidence, but a projection is not a direct observation of the future or the inaccessible past.

The evidence is produced by computation rather than direct experimental manipulation of the target system. Its value depends on transparent assumptions, realistic inputs, sensitivity testing and comparison with independent observations.

What strengthens or limits the finding?

The study covers a large, policy-relevant landscape, propagates uncertainty through 1,000 draws, compares two treatments at three project sizes and publishes its model outputs in a USGS data release. It remains a synthesis of earlier recovery and cost models rather than a randomized field comparison, so its maps estimate relative priorities and should not be treated as guarantees for a particular burned site.

The result explores a plausible explanation or scenario. Its reliability is conditional on assumptions and should be tested against new observations or experiments.

Funding and disclosure context

The recorded funding source is: U.S. Bureau of Land Management and U.S. Geological Survey. The complete conflict-of-interest declaration should be checked in the original publication rather than inferred. Funding or a disclosed relationship does not by itself invalidate a result, but it is relevant when judging design choices, analysis and the need for independent replication.

What it means

The practical lesson is not that one method should replace the other. Managers can use the spatial estimates to identify places where drill seeding's larger expected ecological gain may justify its higher cost, places where lower-cost aerial seeding may stretch a budget further, and places where neither treatment is likely to outperform natural recovery enough to warrant priority. The model also shows why access, topography, burn history and the vegetation present before a fire belong in restoration budgeting rather than being considered only after a method is chosen.

Beyond the abstract

Deeper analysis

More effective and more cost-effective are not synonyms

Drill seeding usually produced the larger modeled vegetation gain, but it also cost more. Cost-effectiveness asks whether the extra improvement is large enough to justify the extra expense. On flatter, accessible ground it often was; on steeper terrain, aerial seeding could provide more expected improvement per dollar even when its absolute ecological effect was smaller.

The small average gains are part of the result

Compared with natural recovery, predicted cover improvement was below one percentage point across most pixels for both methods. That does not make restoration pointless: small changes over an enormous biome can matter, and recovery probability can capture something different from average cover. It does mean the maps are most useful for concentrating effort where returns rise above a generally difficult baseline.

Project scale changes price, not necessarily biology

Median modeled aerial-seeding cost fell from $47 per acre for 100 acres to $13 for 5,000 acres; drill seeding fell from $112 to $44. Those economies of scale raised improvement-per-dollar ratios. The biological model, however, assumed the same per-pixel treatment effect at each project size, so the apparent scale advantage is an economic scenario rather than evidence that larger projects establish sagebrush better.

A prioritization map is a starting point

Regional maps can compare millions of locations consistently, something local trials cannot do. Their weakness is the same generalization. A manager still needs to check soils, seed availability, weather forecasts, invasive plants, equipment limits and objectives on the ground. The strongest test will be whether map-guided decisions improve real outcomes in prospective projects.

Keep the claim in proportion

What it does NOT prove

It does not prove that drill seeding will succeed at 70% or 92% of real restoration sites. Those percentages describe where modeled cost-effectiveness was equal to or greater than aerial seeding under specified scenarios.
It does not show that aerial seeding is ineffective everywhere. Its lower cost made it the better modeled value in some areas, especially as slopes increased.
It does not establish that seeding is always preferable to natural recovery. Predicted gains over no treatment were generally small, and the analysis can help identify low-return locations.
It does not measure every ecological benefit of sagebrush restoration, such as wildlife response, invasive-grass suppression, erosion control or future fire behavior.
It does not provide a site-specific prescription. Local soils, seed mix, weather after treatment, equipment access and management goals require field assessment.

Important limitations

The analysis combined separate published models of ecological recovery and treatment cost; it did not directly compare costs and outcomes within the same randomized restoration projects.
Thirty-year vegetation outcomes followed model projections after a hypothetical wildfire, not a new experiment that tracked every mapped location for three decades.
The ecological model did not estimate an effect of treatment size, so per-pixel effectiveness was assumed to remain constant as modeled project area increased from 100 to 5,000 acres.
Cost data and treatment records were drawn largely from the federal Land Treatment Digital Library and may not represent every contractor, seed market, agency practice or future price.
The analysis considered aerial and drill seeding of Artemisia species. It does not compare the full set of restoration choices, including seedlings, herbicides, invasive-grass control, fuel treatments or different seed mixtures.
Predictions inherit uncertainty and possible bias from the earlier recovery, cost, land-cover, fire-history and soil-moisture datasets used to build them.
Costs were modeled in 2021 dollars and per-acre economies of scale do not imply that a larger project has a lower total cost.
The accessible records did not provide a competing-interests declaration, and changing climate and wildfire regimes may make historical relationships less reliable in the future.

How this fits with previous research

Questions still unanswered

Do projects prioritized with these maps deliver better measured vegetation and wildlife outcomes per dollar than projects selected under current planning methods?
How do unusually wet or dry years after a fire change the relative value of aerial and drill seeding?
Can models include invasive-grass control, seedlings, seed-mixture quality, erosion benefits and avoided future fire costs in the same comparison?
How closely do current local bids, seed prices and equipment costs match estimates expressed in 2021 dollars?
Which locations are better left to natural recovery, and how should uncertainty thresholds influence that decision?

Government verification and context

Relevant U.S. government resources

These resources serve different purposes. A registry can verify what researchers planned, a repository can locate government-funded work, and an agency page can supply authoritative background. None automatically proves that this paper's conclusion is correct.

Government repositoryU.S. Geological Survey

USGS Publications Warehouse ↗

The authoritative catalog of USGS scientific publications, used to check related government research and long-term observational context.

Authoritative contextNational Oceanic and Atmospheric Administration

NOAA research and data ↗

Federal observations and research on climate, oceans, atmosphere and ecosystems provide context for environmental claims. They do not automatically validate a separate model or paper.

Reuse note: Facts and discoveries are summarized here in original language. We link to government material instead of copying it wholesale, and we do not reuse agency logos, photographs, charts or third-party material unless the specific reuse rights are verified.

Sources and provenance