An intermediate layer of standardised data products sitting between primary observations and policy indicators. The definition the family shares:
a minimum set of key and complementary observations on dimensions of Earth systems collected over time that can give a comprehensive yet parsimonious description of the state and trends of the studied system
(Lehman et al. 2020, via Kim et al. 2026, which is the only source behind this page.)
⟨Note: this page carries the mechanics — the families, the class lists, and the two crosswalks transcribed from the source’s landscape tables, which the Markdown conversion destroyed. The argument about what this means for SINE is on the paper page.⟩
The family
| Family | Domain | Steward | Size |
|---|---|---|---|
| ECV — Essential Climate Variables | climate | GCOS, of the World Meteorological Organization | 55 |
| EBV — Essential Biodiversity Variables | biodiversity | GEO BON | 6 classes / 22 subclasses |
| EESV — Essential Ecosystem Service Variables | ecosystem services / Nature’s Contributions to People | proposed by Balvanera et al. 2022 | 6 classes, no formal variable list exists |
| EOV — Essential Ocean Variables | ocean | GOOS, of IOC-UNESCO | — |
ECVs came first (GCOS/WMO) and are the acknowledged model: “Following this success, [GEO BON] established six classes of Essential Biodiversity Variables”. EOVs “in some cases represent building blocks for EBVs” — for the marine environment the basic observation is collected under the EOV framework and used to derive or model the relevant EBV.
⟨Inference: the pattern is a standards genre, not a single standard — each community builds its own essential-variable set against its own steward, and the interlacing between them (an ECV temperature feeding an EBV) is handled case by case rather than by design. That is the seam where the SEEA EA abiotic gap opens.⟩
The EBV hierarchy
Three levels, and the terminology matters because the source is strict about it:
- class (6) — a dimension of biodiversity, e.g. species population, ecosystem structure
- subclass (22), equivalent to a generic EBV variable — e.g. species distribution, ecosystem distribution
- specific EBV variable, or metric — a subclass with taxonomic scope, spatial resolution and temporal resolution fixed, e.g. distribution of Panthera tigris, distribution of tropical rainforests
The selection principle for specific variables: “that they would together be representative of the studied dimension of biodiversity and ecosystems.” The European pilot (EuropaBON) identified and prioritised 84 specific essential variables across terrestrial, freshwater and marine realms (Kissling et al. 2026).
Three of the six classes are measurable at species level (genetic composition, species population, species traits); three are organised at ecosystem level (community composition, ecosystem structure, ecosystem function).
The EESV classes
Six classes corresponding to the stages of the flow of an ecosystem service from nature to society — closely aligned with the levels of the “ecosystem service cascade” model (Haines-Young & Potschin 2010):
| Class | What it is |
|---|---|
| Ecological supply | the ecosystem’s capacity to provide the service |
| Anthropogenic contributions | human contributions to the supply |
| Demand | human need — number of people potentially benefiting, ideally including dependence (costs, substitutes) |
| Use | the amount of the flow actually realised (appropriated) by people |
| Instrumental values | nature’s value as an instrument of human benefit |
| Relational values | principles embedded in or emerging from reciprocal interactions between people and nature |
The source warns explicitly that “supply” and “demand” are borrowed from economics and “do not require traditional economic ‘demand and supply curves’ for their derivation”. A collision follows from this — see the SEEA EA crosswalk below.
EESVs are combined with an ecosystem-service typology (CICES, or the IPBES NCP categories) to name the actual service: pollination, water regulation, recreation. So a complete EESV set is a four-dimensional data cube: spatial × temporal × ecosystem-service-type × EESV class.
EBVs and EESVs are interlinked in both directions: pollinator species population and community composition EBVs drive the ecological supply EESV for pollination; bird species richness (species population EBV) supplies wildlife viewing and so drives a relational value EESV. Use, in turn, affects biodiversity.
Against the IPBES conceptual framework: EBVs inform Nature, EESVs inform Nature’s Contributions to People, and People’s Contributions to Nature is carried by the anthropogenic contribution EESV.
Data cubes, and the three operations
Essential variables are “typically constructed based on sparse observations in space and time”, so they are interpolated into “space-time-biology” data cubes, gaps filled by prediction from auxiliary information and predictive models. “Biology” is the entity dimension: a species for species-focused EBVs, a community or ecosystem for ecosystem-focused ones.
Three operations turn primary data into a reported indicator — “the basic blocks of essential data workflows”:
- Interpolation — sparse points → complete data cube of the same variable
- Modeling (sensu stricto) — assess a new variable from previously assessed variables plus ancillary information
- Aggregation — reduce dimensionality of the final cubes into (composite) indicators requested by policy users
An important role for models is to “represent the causal relationships between the system components (e.g., drivers and biodiversity responses), which allows the detection and attribution of changes through a workflow.”
Known weakness: “there is a lack of understanding about the ‘downstream’ impacts of particular methodological choices in a data workflow context” — for interpolation and aggregation alike.
Storage and semantics: multidimensional cubes in NetCDF (still the format of choice where rich metadata matters — ISIMIP uses it; the Climate Forecast Convention is cited as the model for standardised vocabularies) or cloud-optimised Zarr. Semantic harmonisation via the I-ADOPT framework (Magagna et al. 2022) and the ESM ontology (Affinito et al. 2025). Target property: FAIR. Minimum per record: date and location including altitude or depth, standard units, format described within the data.
Table 1 — the data dimensions each cube must cover
Transcribed from the source (landscape table, re-extracted from the PDF). X = directly required dimension; (x) = auxiliary or indirectly required. Two further dimensions are relevant across all classes (Quoß et al. 2025): metric (different measurements/estimates of the same variable for the same model/data product, potentially in different units) and scenario (different projections of a future state or pathways toward it).
| EBV/EESV class and subclass | Spatial | Temporal | Species | Community, ecosystems | Ecosystem service types | Other relevant dimensions | Example metric |
|---|---|---|---|---|---|---|---|
| Genetic Composition (GC) | |||||||
| Intraspecific genetic diversity | (x) | X | (x) | Populations | allelic richness, heterozygosity | ||
| Genetic differentiation | (x) | X | (x) | Populations | genetic units, distance | ||
| Inbreeding | (x) | X | (x) | Populations | ideal size population | ||
| Effective population size | (x) | X | (x) | Populations | degree of relatedness | ||
| Species Populations (SP) | |||||||
| Species distribution | X | X | X | area of suitable habitat | |||
| Population abundance | X | X | X | estimated counts of individuals | |||
| Species Traits (ST) | |||||||
| Morphology | (x) | (x) | X | trait types | volume, mass, height | ||
| Physiology | (x) | (x) | X | trait types | adaptive capacity | ||
| Phenology | X | X | X | phenological events | timing of colonization or fructification | ||
| Movement | X | X | X | movement types | dispersal ability | ||
| Reproduction | X | X | X | trait types | age at maturity | ||
| Community Composition (CC) | |||||||
| Community abundance | X | X | X | (x) | number of species | ||
| Taxonomic/phylogenetic diversity | X | X | X | (x) | diversity of species | ||
| Trait diversity | X | X | X | (x) | trait types | diversity of traits | |
| Interaction diversity | X | X | X | (x) | interaction types | multitrophic interactions | |
| Ecosystem Structure (ES) | |||||||
| Ecosystem distribution | X | X | X | forest cover | |||
| Live cover fraction | X | X | X | vegetation/canopy layers | vegetation/canopy layers | ||
| Vertical profile | X | X | X | vegetation volume/biomass | |||
| Ecosystem Functions (EF) | |||||||
| Primary productivity | X | X | (x) | X | net primary productivity | ||
| Ecosystem phenology | X | X | (x) | X | phenological events | phytoplankton bloom | |
| Ecosystem disturbances | X | X | (x) | X | disturbance event types | fire, flood, soil erosion, algal bloom | |
| Essential Ecosystem Service Variables | |||||||
| Ecological supply | X | X | (x) | (x) | X | megafauna-based recreational opportunities | |
| Anthropogenic contribution | X | X | (x) | (x) | X | infrastructure to support wildlife viewing | |
| Demand | X | X | (x) | (x) | X | consumer demand for wildlife viewing | |
| Use | X | X | (x) | (x) | X | wildlife watching experiences | |
| Instrumental value | X | X | (x) | (x) | X | revenues of wildlife-based tourism | |
| Relational value | X | X | (x) | (x) | X | value types | stewardship fostered through wildlife viewing |
The crosswalk to SEEA EA (Table 2)
The paper’s own contribution. X = clear direct correspondence; (x) = possible partial or indirect correspondence. SEEA EA condition accounts use the ECT (Ecosystem Condition Typology) classes: A1 physical state, A2 chemical state, B1 compositional state, B2 structural state, B3 functional state, C1 landscape/seascape level.
| EBV/EESV class and subclass | Extent acct. | A1 phys. | A2 chem. | B1 comp. | B2 struct. | B3 funct. | C1 landscape | Services: biophysical flow | Services: monetary flow |
|---|---|---|---|---|---|---|---|---|---|
| Genetic Composition | (x) | ||||||||
| Intraspecific genetic diversity | (x) | ||||||||
| Genetic differentiation | (x) | ||||||||
| Inbreeding | (x) | ||||||||
| Effective population size | (x) | ||||||||
| Species Populations | |||||||||
| Species distribution | X | (x) | |||||||
| Population abundance | X | ||||||||
| Species Traits | |||||||||
| Morphology | (x) | ||||||||
| Physiology | (x) | ||||||||
| Phenology | (x) | ||||||||
| Movement | (x) | ||||||||
| Community Composition | |||||||||
| Community abundance | X | ||||||||
| Taxonomic/phylogenetic diversity | X | ||||||||
| Trait diversity | X | (x) | |||||||
| Interaction diversity | (x) | ||||||||
| Ecosystem Structure | |||||||||
| Ecosystem distribution | X | (x) | |||||||
| Live cover fraction | X | ||||||||
| Vertical profile | X | ||||||||
| Ecosystem Functions | |||||||||
| Primary productivity | X | ||||||||
| Ecosystem phenology | X | ||||||||
| Ecosystem disturbances | X | ||||||||
| Ecosystem Services (EESV) | |||||||||
| Ecological supply | — no dedicated account | ||||||||
| Anthropogenic contribution | — no dedicated account | ||||||||
| Demand | — no dedicated account | ||||||||
| Use | X | ||||||||
| Instrumental value | X | ||||||||
| Relational value | — excluded (exchange values only) |
Three structural findings the table encodes, from the source’s own footnotes:
- The abiotic columns are empty. A1 and A2 “are not covered explicitly in EBV classes (in line with its biotic focus)”, covered only implicitly via ECVs (temperature) and EOVs (nutrients). Landscape-level C1 is likewise not an EBV, though some EBVs (ecosystem distribution) can be used to calculate it. The paper reads the two abiotic ECT classes as “a logical extension of the EBV framework”.
- Species-level EBVs reach the accounts only indirectly. Genetic composition and species traits, and species interactions, “can only indirectly be linked to ECTs”. Species phenology is not considered in SEEA EA condition accounts at all — “lack of clear directionality” — a gap the paper argues against, since phenology shifts under climate change do affect functional state.
- Three EESV classes have no account. Ecological supply, anthropogenic contributions and demand “do not have dedicated ‘accounts’ in the SEEA EA framework”, though all are discussed in SEEA EA Chapter 6 and addressed in recent accounting studies. Relational value is excluded outright: SEEA EA monetary service accounts cover “exclusively use and instrumental values assessed as exchange values.”
The semantic collision. SEEA EA’s supply and use tables are “two different presentations of the same set of numbers (the amount of the service appropriated)” — supply grouped by source ecosystem type, use grouped by user sector. Both therefore map onto EESV use, not onto EESV ecological supply, which is capacity irrespective of use. The structure is “inherited from economic accounting”. Recorded here as a standing translation hazard between the two vocabularies.
Route around the gaps: SEEA EA’s thematic accounting — much more open and flexible than the five regular accounts — can host thematic biodiversity accounts and species accounts, since condition accounts “might not be suitable to host species-level distribution or abundance data for a high number of species”. Species accounts map tightly onto species distribution / species abundance EBVs.
⟨Inference: SEEA EA’s own selection procedure for condition variables “can be seen as a simple operative essential variable framework”, and its reference list of ecosystem services “a selection of services considered ‘essential’ from the perspective of national accounting”. The source says this itself. So there are two essential-variable frameworks in play, built by different communities against different criteria, and the crosswalk is the negotiation between them rather than a lookup table.⟩
The crosswalk to KM-GBF (Figure 3, not transcribed)
EBVs map to Goal A (state of nature), EESVs to Goal B (benefits for society). Worked examples the source names:
- ecosystem distribution EBV → “Extent of Ecosystem by Type” → the “area” component of Goal A
- species and ecosystem distribution EBVs → “Red List Index” → species extinction / status of threatened species, Goal A
- taxonomic/phylogenetic diversity + ecosystem distribution + live cover fraction EBVs → Biodiversity Habitat Index (BHI) and Bioclimatic Ecosystem Resilience Index → Target 1 (spatial planning), Target 2 (ecosystem restoration), Target 8 (climate mitigation and adaptation)
- EESV demand → Target 9 (sustainable use of wild species); EESV supply → Target 11 (NCP restored, maintained, enhanced)
The headline indicator for Goal B is “services provided by ecosystems” and the specific indicators are still under discussion — so the paper’s Goal B examples are illustrative rather than mapped to the actual monitoring framework. Flagged in the figure caption.
The BHI is the best-documented instance of the whole thesis. It assesses how changes in the condition and spatial configuration of natural ecosystems affect the persistence of species diversity in a region, reportable either as average proportion of habitat remaining or as proportion of species expected to persist long-term. Its methodological framework is “purposely designed to allow the indicator to be derived from EBV datasets populated using primary observations from a wide variety of sources” — so the same analytical machinery runs globally, nationally or sub-nationally, with global inputs replaced by local ones where they exist. Instances: San Martin, Peru (Conservation International × Peruvian government, SEEA EA pilot); the Pilbara, Western Australia (cumulative impact of multiple iron-ore mining operations); Republic of Korea (CSIRO × National Institute of Ecology).
⟨Inference: the BHI is the argument that the intermediate layer buys substitutability of inputs — you can improve a national number by swapping in national data without rebuilding the indicator. That is a sharper claim than “harmonisation” and it is the one worth carrying into carbon.⟩
Status
Not a standard. “While EBVs and EESVs have not yet been formally adopted as international standards for ecological monitoring, they are being progressively integrated into data-to-decision workflows.” Applications are voluntary and largely pilots. No formal list of EESV variables exists. Institutionalisation is furthest along in Europe, where the European Commission (2025) has committed to invest in regional coordination of EBV production following the EuropaBON pilot.
Cross-references
- Kim et al. (2026) — the source, and the argument about what this means for SINE
- Nature-related reporting frameworks — KM-GBF, SEEA EA, TNFD, NPI, NGFS
- GEO BON — steward of the EBV framework
- Pathfinder Network — the same intermediate-layer move in carbon, with APIs instead of file formats
- Ledger-based carbon accounting — the competing answer in carbon: not a better intermediate variable but a different unit of account
- Coordination failure vs. bargaining failure
- Circular ecosystem maturity stages — ⟨2026-11-06⟩ the EV layer read as an S3 instance, and the observation that it is financed because it was claimed as a public statistical function (SEEA EA under the UN Statistical Commission) rather than as a market good — the move the circular-economy equivalent never made.
Sources
- Kim et al. (2026) —
sources/whitepapers/2026-09-24-kim-essential-variables-biodiversity-standardization.pdfand its Markdown twin. Tables 1 and 2 above were re-extracted from the PDF directly, not from the Markdown twin: both are landscape-rotated in the original and Marker produced unusable output for them.