Dark Diversity¶
Dark diversity is the set of species absent from a site that belong to its habitat-specific regional species pool.
Core Idea¶
Dark diversity is the set of species that are absent from a study site but belong to its habitat-specific species pool: species in the surrounding region that could potentially inhabit the site's ecological conditions.[1] The site's observed species and its dark diversity together constitute that pool under the chosen definition. A species missing from the site is not automatically dark: it must also be included in the site's potential pool.
The concept makes absence informative. Two sites can have the same observed richness but different pools of suitable regional species, and therefore different proportions of their potential communities represented. Estimating those pools requires assumptions about regional boundaries, ecological suitability, survey scale, dispersal and data quality.[1][2]
How would you explain it like I'm…
The Missing Neighbors
Species That Could Live Here
Absent Members of the Species Pool
Structural Signature¶
- Study site and taxon: choose a spatial and temporal unit and the organisms to assess.
- Regional candidate set: specify the surrounding region and species known there.
- Habitat-specific pool: estimate which regional species could inhabit the site's conditions.
- Observed component: inventory species recorded locally, allowing for imperfect detection.
- Dark component: identify or estimate pool species absent locally.
- Comparison: relate local and dark diversity, with the definition and estimation method stated.[1][2]
Condensed: regional candidates + ecological suitability − local presence → inferred dark diversity.
What It Is Not¶
- Not every species absent from a site. Species outside the specified pool are excluded.
- Not a direct count of undetected organisms. An organism missed by a survey may be present rather than absent; detection error can bias the inferred dark component.
- Not regional diversity or beta diversity. The pool filters the region for the site's ecological conditions.[1]
- Not a guaranteed restoration outcome. A species in an estimated pool is a possible member under the stated conditions, not a promise that it will colonize or persist.
- Not a method-independent number. Different boundaries, data and estimation procedures can produce different pools.[2]
Scope of Application¶
In community ecology, researchers compare observed richness with estimated habitat-specific pools to examine the roles of local and regional constraints. One completeness index is ln(observed richness / dark richness), defined when both counts are positive; another is the observed proportion of the total pool. These related quantities are not interchangeable numerical scales.[2][3] A ratio cannot remove differences caused by inconsistent site definitions or estimation error.
In conservation and restoration, estimated absent but suitable species can suggest questions about habitat condition and dispersal. The estimate must be interpreted alongside field knowledge, imperfect detection, changes in environmental conditions and uncertainty about future establishment.[1]
Worked Examples¶
Grassland thought experiment. In the original formulation, a grassland site has observed species A, B and C. Its regionally assessed grassland pool additionally contains D and E; D and E are the site's dark diversity. Species from other regional habitats outside that pool do not become dark species just because they are absent. This is a theoretical illustration, not a site-specific survey outcome.[1]
Danish plant communities. Fløjgaard and colleagues used co-occurrence-based pool estimation to examine observed and dark vascular-plant diversity across environmental gradients and habitats. Their analysis associated dark diversity with local environmental conditions, organic carbon and competition-related measures; these are study-specific modeled associations, not proof that one intervention will restore particular species.[3]
Practical Use and Limits¶
A defensible estimate reports the site and region, taxonomic scope, sampling effort, pool-estimation procedure and uncertainty. Co-occurrence patterns, habitat preferences and distribution models offer different evidence for pool membership. Incomplete surveys can confuse local presence with absence, and an estimated potential species need not be a feasible colonist under future conditions. When dark richness is zero, the log observed-to-dark index is undefined; report that boundary rather than substituting an arbitrary value.[2]
Neighborhood in Abstraction Space¶
Dark Diversity sits in a sparse region of the domain-specific corpus (92nd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Unclustered & Miscellaneous (2835 abstractions)
Nearest neighbors
- Gamma diversity — 0.81
- Native Species — 0.80
- Species–Area Relationship — 0.78
- Local Range of a Taxon (Teilzone) — 0.78
- Biodiversity Hotspot — 0.78
Computed from structural-signature embeddings · 2026-10-31
References¶
[1] Meelis Pärtel, Robert Szava-Kovats, and Martin Zobel, “Dark diversity: shedding light on absent species,” Trends in Ecology & Evolution 26 (2011): 124–128, doi:10.1016/j.tree.2010.12.004; author manuscript. registry ↩a ↩b ↩c ↩d ↩e ↩f
[2] Rob J. Lewis, Robert Szava-Kovats, and Meelis Pärtel, “Estimating dark diversity and species pools: an empirical assessment of two methods,” Methods in Ecology and Evolution 7 (2016): 104–113, doi:10.1111/2041-210X.12443. registry ↩a ↩b ↩c ↩d ↩e
[3] Camilla Fløjgaard, Jose W. Valdez, Lars Dalby, Jesper Erenskjold Moeslund, Kevin K. Clausen, Rasmus Ejrnæs, Meelis Pärtel, and Ane Kirstine Brunbjerg, “Dark diversity reveals importance of biotic resources and competition for plant diversity across habitats,” Ecology and Evolution 10 (2020): 6078–6088, doi:10.1002/ece3.6351. registry ↩a ↩b