Skip to content
Ecological Intelligence

Ecological evidence.
Independent perspective.
Explore Wild Lands ↗

Ecological Intelligence / Subject

Freshwater

North America

River mussel communities are sorting along flow regimes more consistently than along map scale

Evidence
Peer-reviewed multi-scale river field analysis
Geography
Colorado River basin, Texas, USA
System
Rivers / freshwater mussel habitat

WHAT THE EVIDENCE SAYS

Researchers paired freshwater-mussel abundance from 480 sites with long-term hydrologic records and tested patterns across sub-basins, tributaries and mainstems. Three flow gradients repeatedly structured life-history composition: flow stability, flow variability and low-flow disturbance. Equilibrium species were associated with sustained stable flows, opportunistic species with zero-flow disturbance, and periodic species with greater flow variability.

WHY IT MATTERS / OUR READ

For managers, mean discharge is not the whole habitat. If long-lived mussels are a conservation objective, monitoring should preserve the flow components that organisms actually experience—summer-flow persistence, zero-flow days, pulse duration and variability—and use those data alongside population surveys when evaluating water operations or drought response.

CAVEATS / UNCERTAINTY

This is an observational analysis from one Texas basin and does not establish a universal environmental-flow prescription. Host fish, substrate, water quality and connectivity also affect mussels, and life-history guilds summarize rather than replace species-specific requirements.

SOURCE

Freshwater Biology — primary study
View original evidence ↗

Across 480 Texas Colorado River sites, long-lived equilibrium species tracked stable flows, opportunists tracked zero-flow disturbance, and periodic species increased with flow variability.

Read the Wild Lands assessment →
Peer-reviewed Latin America & Caribbean

Chocó forest regeneration restored riparian herp assemblages imperfectly—and not monotonically

Evidence
Peer-reviewed
Geography
Canandé Reserve, Esmeraldas Province, Ecuador
System
Riparian forest streams

WHAT THE EVIDENCE SAYS

Standardized surveys at 17 stream transects compared primary forest, regenerating secondary forest on former cacao/pasture land, and actively used agricultural areas. Active sites held the lowest amphibian and reptile diversity. Regenerating sites sometimes showed higher observed richness and diversity than primary forest, but those contrasts were not statistically significant. Canopy and understory structure were important predictors of amphibian assemblages, while reptile patterns reflected broader site conditions.

WHY IT MATTERS / OUR READ

Recovery does not have to recreate a simple monotonic march toward an old-forest richness value. Structural habitat conditions can recover at different rates, and community composition may be more informative than whether one diversity index has returned.

CAVEATS / UNCERTAINTY

Seventeen stream sites provide a useful field gradient but limited replication for broad causal claims. Richness differences between regenerating and primary forest were not statistically significant, and land-use history, stream condition and detectability can all contribute to assemblage differences.

SOURCE

Biotropica — primary study
View original evidence ↗

Across 17 Ecuadorian stream sites, active agricultural reaches were the clearest low point; regenerating sites sometimes matched or exceeded old forest in richness, but those differences were not statistically decisive.

Read the Wild Lands assessment →
Peer-reviewed North America

Low-dose copper suppressed zebra-mussel recruitment for nearly two years—but the non-target response was not zero

Evidence
Peer-reviewed
Geography
Lake Minnetonka, Minnesota, USA
System
Freshwater lake / invasive mussel management

WHAT THE EVIDENCE SAYS

In a 66.3-ha bay, a 10-day low-dose copper treatment reduced zebra-mussel veliger density from 6.0 to 0.3 per liter and produced settlement roughly 1,900 times lower than an untreated bay immediately after treatment; recruitment remained suppressed for nearly two years, while caged adult mortality was about 30%. A companion non-target study found zooplankton abundance and diversity initially declined in the treated bay but returned to pretreatment/reference levels within a year, no significant between-bay effect on benthic invertebrate abundance or diversity was detected overall, and native mussels accumulated more copper than fish.

WHY IT MATTERS / OUR READ

For managers, the useful result is not ‘copper works.’ It is that a lower-dose suppression strategy can change recruitment without complete adult eradication, while imposing measurable but partly transient non-target effects. Whether that trade is acceptable depends on the management objective, permitting context, native-mussel exposure, monitoring capacity and the ecological cost of leaving the invader untreated.

CAVEATS / UNCERTAINTY

This was one bay-level treatment with a reference bay in one Minnesota lake, and target and non-target studies report nominal versus measured treatment concentrations differently. Fish inference was limited by low cage survival, non-target communities were dynamic, and results should not be generalized to other water chemistry, taxa, formulations or treatment goals without site-specific regulatory and ecological review.

SOURCE

USGS / Environmental Management — target effects
View original evidence ↗

Companion primary study — non-target effects
View supporting source ↗

A Lake Minnetonka field treatment sharply reduced veligers and settlement; a companion study found short-term zooplankton declines, mussel copper uptake and other transient effects that managers have to weigh against the control objective.

Read the Wild Lands assessment →
Peer-reviewed Europe

Large pike connected littoral and pelagic food webs by changing diets through life

Evidence
Peer-reviewed
Geography
Lake Bourget, France
System
Large lake littoral and pelagic food webs

WHAT THE EVIDENCE SAYS

Sequential stable-isotope measurements from eye lenses of 17 large pike reconstructed lifetime resource use. Every fish showed a major littoral-to-pelagic dietary shift, but the timing and extent differed among individuals; median lifetime littoral reliance was 52.5%.

WHY IT MATTERS / OUR READ

Food-web coupling can be generated by ontogeny rather than by a species occupying the same trophic role throughout life. Population models that average individuals may hide when and how top predators move energy between habitats.

CAVEATS / UNCERTAINTY

The study is retrospective and based on 17 large pike from one lake. Eye-lens isotope trajectories infer resource pathways rather than directly tracking movement, and results should not be generalized to all pike populations or lake types.

SOURCE

Freshwater Biology
View original evidence ↗

EvidencePeer-reviewedGeographyLake Bourget, FranceSystemLarge lake littoral and pelagic food webs WHAT THE EVIDENCE SAYS Sequential stable-isotope measurements from eye lenses of 17 large pike reconstructed lifetime resource use. Every

Read the Wild Lands assessment →
Peer-reviewed Global

A 3,400-metagenome atlas brings freshwater microeukaryotes into species-resolved view

Evidence
Peer-reviewed
Geography
362 lakes across seven continents
System
Freshwater lakes / microbial eukaryotes

WHAT THE EVIDENCE SAYS

Researchers assembled a global mitochondrial genomic atlas from more than 3,400 freshwater metagenomes spanning 362 lakes on seven continents, recovering roughly 20,000 mitochondrial metagenome-assembled genomes and exposing substantial previously unresolved microeukaryotic diversity.

WHY IT MATTERS / OUR READ

The dataset expands the genomic reference base available for freshwater monitoring and ecological inference, especially for organisms that routine morphology or sparse barcode libraries can miss.

CAVEATS / UNCERTAINTY

Sampling remains uneven across geography and taxa, the operational species threshold is methodological rather than a complete taxonomic census, and a large atlas is not equivalent to globally complete biodiversity coverage.

SOURCE

Nature Communications
View original evidence ↗

EvidencePeer-reviewedGeography362 lakes across seven continentsSystemFreshwater lakes / microbial eukaryotes WHAT THE EVIDENCE SAYS Researchers assembled a global mitochondrial genomic atlas from more than 3,400 freshwater metagenomes spanning 362 lakes

Read the Wild Lands assessment →
Preprint Global

Human pressure did not produce a universal freshwater homogenization signal

Evidence
Preprint
Geography
Global freshwater studies
System
Freshwater assemblages

WHAT THE EVIDENCE SAYS

A global preprint meta-analysis of 495 reference–impact contrasts found higher beta diversity in impacted assemblages overall, indicating differentiation rather than universal homogenization across taxonomic and functional facets. Yet presence–absence and quantitative data pointed in opposite directions, and removing the heavily represented fish and invertebrate studies reversed the pooled response toward homogenization.

WHY IT MATTERS / OUR READ

The popular shorthand that human disturbance makes ecological communities everywhere more alike is too simple. The inferred direction depends on what organisms are represented, whether abundance matters, and how reference and impacted sites are arranged.

CAVEATS / UNCERTAINTY

This is a version-1 preprint and has not been peer reviewed. The pooled result is highly context-dependent, study designs and taxa are unevenly represented, and the authors explicitly show that reasonable analytical subsets can reverse the headline direction.

SOURCE

EcoEvoRxiv
View original evidence ↗

A 495-contrast preprint meta-analysis found overall biotic differentiation, but the direction flipped with abundance data and taxonomic representation.

Read the Wild Lands assessment →
Peer-reviewed Africa

Around a Nigerian hydropower reservoir, semi-natural patches held more bird diversity than the disturbed edges

Evidence
Peer-reviewed
Geography
Shiroro Reservoir, Niger State, Nigeria
System
Hydropower reservoir / savanna landscape

WHAT THE EVIDENCE SAYS

Surveys around seven Shiroro Reservoir sites recorded 21,418 individual birds representing 179 species. Semi-natural habitats supported higher taxonomic and functional diversity than more disturbed or infrastructure-proximate locations; broad species-richness and Shannon responses were not strongly seasonal, although some functional metrics differed between wet and dry periods.

WHY IT MATTERS / OUR READ

Large infrastructure landscapes are not ecologically uniform. Retained vegetation structure and semi-natural habitat around reservoir margins can matter for community composition even where the reservoir itself dominates the map.

CAVEATS / UNCERTAINTY

This is observational evidence from one reservoir landscape, not a before–after hydropower experiment. Site differences are correlated with several habitat and disturbance factors, and measured vegetation variables explained only part of the variation.

SOURCE

Frontiers in Ecology and Evolution
View original evidence ↗

Across seven Shiroro Reservoir sites, taxonomic and functional diversity varied more by location and disturbance than by broad wet–dry season counts.

Read the Wild Lands assessment →
Peer-reviewed Global

Aquatic fungi just got a global dataset—and its gaps are as informative as its scale

Evidence
Peer-reviewed
Geography
Global — 1,452 aquatic sites in 87 countries
System
Freshwater and marine aquatic ecosystems

WHAT THE EVIDENCE SAYS

The FunAqua data descriptor combines 1,076 sediment and 1,299 filtered-water samples from 1,452 sites in 87 countries and reports 50,152 fungal OTUs using long-read ITS metabarcoding. European samples make up about 56% of the dataset, with freshwater lakes and rivers much more heavily represented than several coastal and specialized aquatic habitats.

WHY IT MATTERS / OUR READ

This is valuable as both a biodiversity resource and a map of where the evidence is thin. Global microbial baselines become more useful when sampling geography, biome coverage, sequencing depth and laboratory differences remain visible rather than disappearing behind the word 'global.'

CAVEATS / UNCERTAINTY

Sampling is geographically and environmentally uneven, protocols and preservation histories vary among contributing campaigns, ITS barcoding has taxonomic limitations, and recovered DNA does not by itself distinguish actively growing aquatic fungi from spores or material transported into the sample.

SOURCE

Scientific Data
View original evidence ↗

FunAqua assembles long-read metabarcoding from 1,452 sites in 87 countries, but Europe supplies more than half the samples and fungal sequencing depth remains uneven.

Read the Wild Lands assessment →
Peer-reviewed Europe

Lake Geneva’s sediment record is being written by river intrusions below the surface

Evidence
Peer-reviewed
Geography
Lake Geneva / Rhône River, Switzerland and France
System
Large lake / river-lake sediment transport

WHAT THE EVIDENCE SAYS

Sediment traps at 10, 30, 50 and 100 m from 2020–2024 showed the Rhône as the primary driver of Lake Geneva sedimentation, with density-driven intrusions carrying particles below the thermocline. Floods generated the largest particle fluxes, while drought sharply reduced sediment supply and storms redistributed stored material.

WHY IT MATTERS / OUR READ

Monitoring only surface water or annual sediment totals can hide the transport pathway. Stratification and event type determine where river-borne material goes inside a large lake.

CAVEATS / UNCERTAINTY

This is one large peri-alpine lake observed from a fixed platform and its associated river network. The study resolves transport well but does not make its exact depth patterns universal to other lakes.

SOURCE

Aquatic Sciences
View original evidence ↗

Five years of depth-resolved traps show why river inflow, storms and stratification need to be monitored together rather than treating lake sedimentation as a surface process.

Read the Wild Lands assessment →
Peer-reviewed Africa

One invasive fish can suppress another—and leave a native remnant in the middle

Evidence
Peer-reviewed
Geography
Breede River, Western Cape, South Africa
System
River fish community

WHAT THE EVIDENCE SAYS

A river survey recorded 2,546 fish representing 17 species, seven of them introduced. Sharptooth catfish frequency was negatively associated with fish diversity, while invasive black bass showed the opposite broad association; flow barriers, high catfish abundance and turbidity appeared to suppress visually hunting bass enough for endangered Berg-Breede whitefish to persist at very low frequency.

WHY IT MATTERS / OUR READ

In invaded systems, management can face interacting pressures rather than one clean native-versus-invasive axis. Removing one stressor can change the effect of another, which makes food-web and barrier context relevant to control decisions.

CAVEATS / UNCERTAINTY

The whitefish remnant is persistence, not recovery. The evidence is observational and site-specific, and antagonism among invasive fishes is not an argument to retain invasive species or barriers; intervention effects would need direct testing.

SOURCE

Biological Invasions
View original evidence ↗

In South Africa’s Breede River, flow barriers, turbidity and abundant sharptooth catfish appear to damp black-bass predation enough for an endangered whitefish to persist at extremely low frequency.

Read the Wild Lands assessment →
Peer-reviewed Latin America & Caribbean

A forest fragment can help a stream even when the rest of the reach is farm country

Evidence
Peer-reviewed
Geography
Espírito Santo, Brazil
System
Headwater agricultural streams / Atlantic Forest

WHAT THE EVIDENCE SAYS

Across 16 reaches in eight agricultural catchments, upstream forest fragments reduced sediment, moderated temperature and increased substrate diversity downstream, while downstream forest reaches supported recovery in invertebrate community attributes. Leaf-litter experiments also showed context-dependent effects on decomposition.

WHY IT MATTERS / OUR READ

Riparian protection can have effects that propagate beyond the mapped forest polygon. Placement along the flow network matters as much as simply counting wooded reach length.

CAVEATS / UNCERTAINTY

Only eight catchments and four replicates of each reach configuration were studied, so causal generalization is limited. The results are from one Atlantic Forest agricultural landscape and do not quantify every water-quality or biodiversity outcome.

SOURCE

Oecologia
View original evidence ↗

Eight Atlantic Forest catchments show upstream and downstream forest fragments buffering different parts of the ecological response to agricultural riparian loss.

Read the Wild Lands assessment →
Peer-reviewed Arctic

Alaska’s orange rivers now have a downstream chemistry record, not just a color signal

Evidence
Peer-reviewed
Geography
Brooks Range, northwest Arctic Alaska, USA
System
Permafrost-fed river networks

WHAT THE EVIDENCE SAYS

Researchers sampled six affected northwest Arctic Alaska watersheds during 2022–2024 and reconstructed a regional onset around 2019. Acidic, metal-rich tributaries entered larger rivers where acidity could be buffered while elevated metal concentrations persisted far downstream, in some cases for tens of miles.

WHY IT MATTERS / OUR READ

The important monitoring lesson is that visual orange staining and local pH do not describe the whole exposure pathway. A receiving river can buffer acidity while still transporting mobilized metals through the network.

CAVEATS / UNCERTAINTY

This is an emerging, observational river-chemistry phenomenon linked to thawing permafrost and weather history, not a controlled attribution experiment. The authors report no known current human or wildlife harm and no massive fish die-offs; ecological and drinking-water consequences remain under study.

SOURCE

AGU Advances
View original evidence ↗

UC Davis research summary
View research summary ↗

Across six Brooks Range watersheds, acid-rock-drainage chemistry showed abrupt onset around 2019 and metal enrichment persisted far downstream even where mainstem pH was buffered.

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 →
Peer-reviewed Africa

South Africa’s invasive Tarebia snails point to one globally successful lineage—but one marker cannot reconstruct the invasion

Evidence
Peer-reviewed
Geography
South Africa, with comparison to the species’ global invasive range
System
Freshwater invasive-snail populations and parasite-host networks

WHAT THE EVIDENCE SAYS

Genetic analysis of invasive Tarebia granifera in South Africa found the sampled haplotypes nested within a major invasive lineage also represented elsewhere in the species’ introduced range. The low diversity observed in South Africa is consistent with a limited founding history followed by expansion, while the species’ role as an intermediate host for trematodes adds a disease-ecology dimension to its spread.

WHY IT MATTERS / OUR READ

For invasion surveillance, lineage information can help distinguish repeated introductions from expansion after establishment and can identify whether pathogen-host concerns are traveling with a widespread invasive lineage.

CAVEATS / UNCERTAINTY

The study relies heavily on a single mitochondrial marker, which is useful for broad lineage assignment but weak for fine-scale reconstruction of introduction number, source population or dispersal route. Disease risk was not measured as a direct outbreak outcome.

SOURCE

Aquatic Sciences
View original evidence ↗

Mitochondrial evidence places South African Tarebia granifera within the same major invasive lineage found elsewhere, consistent with rapid expansion from limited introductions.

Read the Wild Lands assessment →

Keep exploring

Follow the systems. Find the connections.

WildlifeLandWaterOceansClimateCitiesOne HealthGovernanceMethods / Monitoring
Browse the archive →