Skip to content
Ecological Intelligence

Ecological evidence.
Independent perspective.
Explore Wild Lands ↗

Ecological Intelligence / Subject

GIS

Peer-reviewed North America

A restored floodplain can regrow trees without regaining its original small-scale terrain

Evidence
Peer-reviewed
Geography
Sacramento River floodplain, California, United States
System
Riparian forest / floodplain restoration / LiDAR

WHAT THE EVIDENCE SAYS

Comparing restored and remnant riparian-forest patches with fine- and coarse-resolution LiDAR, researchers found 2-meter terrain ruggedness 76% lower in restored forest and only gradually recovering. After 11 years, remnant forests retained 13% more foliage-height diversity and 23% greater canopy rugosity. By about 20 years, vegetation structural heterogeneity was similar in restored and remnant forests, while fine microtopographic differences persisted.

WHY IT MATTERS / OUR READ

For projects starting on graded or simplified floodplains, decide whether seedling establishment, canopy complexity or terrain heterogeneity is the actual goal. Where hydrology and permitting allow, build microtopographic variation into the initial design and monitor it alongside vegetation using repeatable terrain metrics.

CAVEATS / UNCERTAINTY

These are relationships within one dam-influenced river system, not a controlled test that creating bumps will guarantee better wildlife outcomes. A LiDAR proxy measures structure, not all biotic function; any earthwork must be checked against flood conveyance, native communities and site constraints.

SOURCE

Restoration Ecology — Constantz et al. (2026)
View original evidence ↗

Sacramento River LiDAR observations show that vegetation heterogeneity and floodplain microtopography recover on different clocks.

Read the Wild Lands assessment →
Peer-reviewed Europe

Rome's river was a corridor for some dispersers and a barrier for others

Evidence
Peer-reviewed
Geography
Rome metropolitan region, Italy
System
Urban protected areas / corridor modeling

WHAT THE EVIDENCE SAYS

Researchers modeled protected-area connectivity with circuit and graph theory under nine resistance scenarios representing three dispersal modes. The western and peripheral city sectors were generally more permeable than the center and east. The Tiber River appeared as either connective space or resistance depending on the assumed species' movement strategy. Shared candidate links often followed rivers and urban green spaces.

WHY IT MATTERS / OUR READ

Urban landscape plans should test which animals can actually cross a proposed corridor; a riverbank or park is not universally connected simply because it is green in an aerial image.

CAVEATS / UNCERTAINTY

Results are based on modeled resistance and dispersal assumptions, not observed migration or survival. Rome's urban form and species assemblages limit direct transfer of the identified corridor locations to other cities.

SOURCE

Landscape Ecology — Russo et al. (2026)
View original evidence ↗

Urban connectivity models suggest that no single green-link map represents all wildlife movement.

Read the Wild Lands assessment →
Peer-reviewed Europe

Tiny field-edge strips changed where pollinator models placed habitat

Evidence
Peer-reviewed
Geography
Three agricultural landscapes, Netherlands
System
Agricultural landscape / pollinator habitat

WHAT THE EVIDENCE SAYS

Fine-resolution (10-meter) species-distribution models incorporated herbaceous linear landscape elements as cover or distance covariates. Compared with models lacking these features, inclusion improved predictions and reduced overprediction in fields and underprediction at edges; improvements were evaluated against independently collected field records as well as withheld training data.

WHY IT MATTERS / OUR READ

Restoration opportunity maps need to capture narrow field margins and ditches, not just broad land-cover classes; independent field validation matters when models guide scarce conservation funds.

CAVEATS / UNCERTAINTY

Model improvement is not proof that newly adding a strip will cause equivalent abundance gains. Transferability from three Dutch landscapes to other crops, climates or taxa remains to be tested.

SOURCE

Landscape Ecology — Moens et al. (2026)
View original evidence ↗

Dutch bee and hoverfly models improved when linear habitat elements were mapped instead of being averaged into cropland.

Read the Wild Lands assessment →
Peer-reviewed Asia

Kerala's multi-hazard risk ranking survived an uncertainty test

Evidence
Peer-reviewed
Geography
Pampa River basin, Kerala, India
System
Watershed flood, drought and landslide hazard

WHAT THE EVIDENCE SAYS

Researchers combined 30-metre terrain metrics, expert-weighted criteria and 10,000 Monte Carlo iterations to prioritize 31 sub-watersheds. Several high-risk units ranked consistently, with household-level damage data broadly supporting model priorities.

WHY IT MATTERS / OUR READ

Ranking stability should precede investment, but hazard susceptibility is not interchangeable with exposure, institutional capacity or mitigation performance.

CAVEATS / UNCERTAINTY

Expert-weight perturbations were limited to plus or minus ten percent; resulting map classes and validation apply to the study basin.

SOURCE

Natural Hazards — Sekar et al. (2026)
View original evidence ↗

An Indian watershed tool evaluated 31 units against damage reports.

Read the Wild Lands assessment →
Peer-reviewed North America

Maps of wildlife movement and climate refugia did not always agree

Evidence
Peer-reviewed
Geography
Far western North America; 103 conservation connectivity maps
System
Habitat connectivity; climate adaptation; spatial decision-making

WHAT THE EVIDENCE SAYS

A comparison of 103 connectivity and climate-adaptation maps in far western North America documented divergent priorities among approaches that represented existing ecological movement patterns versus longer-term climate-related connectivity. Areas of convergence were fewer and more selective than a single generalized corridor map might suggest.

WHY IT MATTERS / OUR READ

GIS synthesis should show where layers agree, where they conflict and which biological decision each serves. A route useful for present-day movement may not be the same landscape investment needed under future climate change.

CAVEATS / UNCERTAINTY

A comparison of modeled spatial products is not a direct test of animal movement or realized adaptation; input scales and assumptions influence apparent agreement.

SOURCE

Landscape Ecology · western North America connectivity-map comparison
View original evidence ↗

Comparing 103 connectivity maps reveals priorities that depend heavily on model purpose.

Read the Wild Lands assessment →
Commentary Asia

Wildlife cameras may be an underused understory phenology sensor

Evidence
Commentary
Geography
Northeast China example; broader forest-monitoring applications
System
Forest understory seasonal development; phenology

WHAT THE EVIDENCE SAYS

A Nature Reviews Earth & Environment commentary describes camera-based understory monitoring and an example integrating camera and drone information that identified asynchronous seasonal patterns across forest height strata. It proposes broader use to study trophic timing and climate-related shifts.

WHY IT MATTERS / OUR READ

Camera networks installed for mammals may also record vegetation change invisible to canopy-level remote sensing, potentially extending monitoring value at low marginal field cost.

CAVEATS / UNCERTAINTY

This is a methods perspective, not a newly validated global phenology dataset or proof that all camera placements permit unbiased time-series measurement.

SOURCE

Nature Reviews Earth & Environment — methods commentary
View original evidence ↗

A methods commentary points to camera records as a complement to satellite and drone views of vegetation timing beneath forest canopies.

Read the Wild Lands assessment →
Peer-reviewed North America

Fire researchers linked airborne lasers, ground scans and field plots in one fuel dataset

Evidence
Peer-reviewed
Geography
Southeastern and western United States
System
Wildland fuels; pine stands and grasslands

WHAT THE EVIDENCE SAYS

An integrated sampling program paired airborne LiDAR, terrestrial laser scans, structure-from-motion imagery and destructive fuel measurements at 18 sites. The co-located fuel objects and metrics are available in a research data repository for training and evaluating 3-D fire-behavior models.

WHY IT MATTERS / OUR READ

Structural mapping becomes more decision-useful when remote measurements are calibrated against actual fuel objects and concentrations on the ground.

CAVEATS / UNCERTAINTY

This paper describes a datasets-and-methods foundation, not proof that resulting fire forecasts are now accurate everywhere. Operational transfer requires model training and field checks.

SOURCE

Fire Ecology / USDA Forest Service — primary data-method paper
View original evidence ↗

Nine southeastern pine, five western pine and four western grassland sites now have aligned 3-D fuel data to test fire-model inputs.

Read the Wild Lands assessment →
Peer-reviewed Asia

One coastal resilience map was not enough for China’s local bays

Evidence
Peer-reviewed
Geography
Mainland China and Rongcheng, Shandong, China
System
Coastal social-ecological land–sea system

WHAT THE EVIDENCE SAYS

A three-tier, ten-attribute spatial framework combined nominal-2025 national indicators with a separately weighted local case study. National means were 0.359 for coastal land and 0.470 for coastal waters; more detailed Rongcheng maps exposed within-bay variability, including aquaculture-influenced areas.

WHY IT MATTERS / OUR READ

A national index can screen for places that need attention, but monitoring and management limits must be set at the scale of local ecological and institutional processes.

CAVEATS / UNCERTAINTY

These are model-derived composite capacity scores, not verified post-storm recovery or a causal test. National and local values cannot be compared directly because their indicators and weightings differ.

SOURCE

Frontiers in Ecology and Evolution — primary coastal index analysis
View original evidence ↗

A national land-and-water index and a nested Rongcheng diagnosis differed in what they could resolve; scores measured capacity, not observed recovery.

Read the Wild Lands assessment →
Latin America & Caribbean

A Colombian watershed model maps flood and landslide risk with green infrastructure options

Evidence
Peer-reviewed spatial planning model
Geography
Guarapas River catchment, Huila, Colombia
System
Tropical catchment / blue-green infrastructure

WHAT THE EVIDENCE SAYS

Weighted multicriteria overlays identify more than 60% of the catchment as suitable for at least one blue-green measure; historical hazard records were used for validation. The paper's up-to-55% risk-reduction claim is from modelled scenarios, not deployed projects.

WHY IT MATTERS / OUR READ

This is an approach for targeting capital in data-poor catchments, but the next test is field hydrology, maintenance and institutional feasibility.

CAVEATS / UNCERTAINTY

Risk reduction is simulated and depends on scoring weights and input resolution, not measured performance after construction.

SOURCE

Sustainability and Climate Change · original spatial decision model
View original evidence ↗

A planning framework links hazard records with places for forest, floodplain, pond and wetland investments.

Read the Wild Lands assessment →

Bohai Bay's benthic impacts are not confined to direct development footprints

Evidence
Peer-reviewed cumulative impact model
Geography
Bohai Bay, China
System
Subtidal marine sediments / benthic life

WHAT THE EVIDENCE SAYS

Spatial models and benthic taxa show the infralittoral mud zone experiencing extensive eutrophication risk. The authors report about 32% of its area affected by environmental pressure compared with 4.72% physical damage and 0.17% direct loss.

WHY IT MATTERS / OUR READ

Project-by-project footprint accounting can miss dominant diffuse stressors in busy bays; habitat-specific sensitivity maps may sharpen monitoring priorities.

CAVEATS / UNCERTAINTY

Mapped indices depend on weights, input quality and sensitivity assumptions; co-occurring stress and tolerant taxa cannot establish single-driver causation.

SOURCE

Frontiers in Marine Science · spatial impact study
View original evidence ↗

A GIS pressure–sensitivity assessment traces seabed stress from nutrients, physical damage and habitat removal.

Read the Wild Lands assessment →
Peer-reviewed Europe

At 30 meters, Czech modeled fire behavior looked much more severe than at 100 meters

Evidence
Peer-reviewed
Geography
Czech Republic
System
Forest and agricultural wildfire landscape

WHAT THE EVIDENCE SAYS

A national 30-m fuel classification produced markedly more modeled high-rate spread, long flame lengths and high fireline intensity than a 100-m European fuel product under otherwise identical FlamMap conditions. The finer map also classified much less area as nonburnable.

WHY IT MATTERS / OUR READ

Fuel-map resolution can change modeled hazard enough to alter where treatment, preparedness or infrastructure decisions appear most urgent.

CAVEATS / UNCERTAINTY

These are model outputs conditioned on different fuel inputs, not observed fire behavior. The authors note that the new classification still needs validation against fire observations before it is treated as an operational truth layer.

SOURCE

Fire Ecology
View original evidence ↗

EvidencePeer-reviewedGeographyCzech RepublicSystemForest and agricultural wildfire landscape WHAT THE EVIDENCE SAYS A national 30-m fuel classification produced markedly more modeled high-rate spread, long flame lengths and high fireline

Read the Wild Lands assessment →
Peer-reviewed

Seasonal livestock maps can separate where grazing pressure actually occurs

Evidence
Peer-reviewed
Geography
China grazing-production systems
System
Rangelands / pastoral systems

WHAT THE EVIDENCE SAYS

A new dataset reconstructs 1-km warm- and cold-season distributions of cattle and sheep/goats across China from 2000–2021 using a stacking ensemble. Internal test performance was high and external city-level validation was lower but still useful, providing a spatially explicit alternative to annual aggregate livestock counts.

WHY IT MATTERS / OUR READ

Seasonal maps can make grazing exposure more ecologically interpretable than administrative totals, especially where animals move between summer and winter ranges.

CAVEATS / UNCERTAINTY

These are modeled distributions, not direct animal counts in every grid cell. External validation is weaker than internal performance, and livestock density alone does not establish ecological impact without vegetation, climate and management context.

SOURCE

Scientific Data
View original evidence ↗

EvidencePeer-reviewedGeographyChina grazing-production systemsSystemRangelands / pastoral systems WHAT THE EVIDENCE SAYS A new dataset reconstructs 1-km warm- and cold-season distributions of cattle and sheep/goats across China from

Read the Wild Lands assessment →
Peer-reviewed

Change the map scale and Hangzhou Bay’s ecological trend changes with it

Evidence
Peer-reviewed
Geography
Hangzhou Bay Urban Agglomeration, China
System
Urban / coastal metropolitan landscape

WHAT THE EVIDENCE SAYS

A Google Earth Engine analysis tracked a remote-sensing ecological index from 2000–2024 at 1-, 3-, 5- and 10-km grids and county scale. The regional RSEI increased by about 0.0423 overall, but the share classified as improving rose from 75.34% at 1 km to 84.25% at county scale, demonstrating how aggregation can hide opposing local trajectories. Land-use intensity was the strongest SHAP contributor in the explanatory model.

WHY IT MATTERS / OUR READ

For metropolitan ecological planning, scale is part of the result. County or city averages can be useful for screening, but site and corridor decisions need nested spatial analysis so local degradation is not erased by improvement elsewhere.

CAVEATS / UNCERTAINTY

RSEI combines greenness, wetness, dryness and surface temperature; it is not a direct biodiversity or ecosystem-integrity measure. The study is observational and cannot assign causal effects to land use or policy, and its machine-learning validation was not designed as a strict spatial or temporal holdout test.

SOURCE

Frontiers in Ecology and Evolution
View original evidence ↗

A 25-year remote-sensing analysis found coarse administrative units made improvement look more widespread while hiding local opposing trends along urban corridors and the bay.

Read the Wild Lands assessment →
Peer-reviewed

For migratory shorebirds, the landscape around a habitat patch improved the prediction

Evidence
Peer-reviewed
Geography
Coastal East Asian–Australasian Flyway
System
Coastal wetlands / migratory shorebird habitat

WHAT THE EVIDENCE SAYS

Across six migratory shorebird species in the non-breeding season, species-distribution models that incorporated neighborhood habitat composition and configuration generally predicted habitat suitability better than conventional pixel-scale or distance-only models. Important predictors included distance to tidal flats and seasonal inland wetlands, tidal-flat area and patch density, land-cover diversity, and elevation; predicted suitability was only weakly positively correlated with observed abundance for most species.

WHY IT MATTERS / OUR READ

Habitat is not only the patch under a bird. For mobile species in heavily altered coastal landscapes, surrounding composition and configuration can change which places function as habitat and can change conservation-priority maps.

CAVEATS / UNCERTAINTY

This is modeling evidence for six species, not a demographic test of survival or population growth. Abundance correlations were weak, and the publisher is currently sharing an accepted early version that remains subject to copyediting.

SOURCE

Communications Earth & Environment
View original evidence ↗

Across six East Asian–Australasian Flyway species, models that added tidal-flat amount, patch density and surrounding land-cover diversity generally outperformed pixel- or distance-only habitat models.

Read the Wild Lands assessment →
Peer-reviewed North America

A U.S. freshwater conservation map changes when resilience and connectivity are treated as properties of the network

Evidence
Peer-reviewed
Geography
Conterminous United States
System
Rivers, streams and small watersheds

WHAT THE EVIDENCE SAYS

The study scored HUC-12 watersheds for connectivity, condition, water availability and naturalness, then overlaid 91 freshwater biodiversity assessments. The resulting network covers 35% of the conterminous U.S.: 19% scored resilient and 16% restorable. It contains 67% of recognized biodiversity areas and 75% of free-flowing rivers; just 6% is both protected and resilient, while 14% is unprotected and restorable.

WHY IT MATTERS / OUR READ

The useful contribution is a decision surface rather than a single protection target. It separates places that are already resilient from places where restoration could change network function, making connectivity and climate context explicit in freshwater prioritization.

CAVEATS / UNCERTAINTY

This is a spatial prioritization framework, not evidence that conserving exactly 35% guarantees biodiversity persistence. Human water needs and many local implementation constraints are not embedded in the core network and must be layered into actual decisions.

SOURCE

PLOS ONE
View original evidence ↗

A continental planning analysis identifies a 35% freshwater network that combines representation, resilience, connectivity and biodiversity—but only 6% is both protected and resilient.

Read the Wild Lands assessment →

Keep exploring

Follow the systems. Find the connections.

WildlifeLandWaterOceansClimateCitiesOne HealthGovernanceMethods / Monitoring
Browse the archive →