Curated Species nodes from the nine agnt eco agents, ranked by a six-component centrality score derived from the knowledge graph itself — not from how often the curator typed "important".
Importance in an ecosystem is multi-dimensional. A single
number is a compromise — but a careful compromise is more useful than
narrative pull alone. Each species gets six component scores, each
normalised inside its agent's set, then combined into one
0–100 figure.
A weighted sum of all the species' edges in the graph. Trophic
edges (FEEDS, MYCORRHIZAL_WITH,
DECAYS) count more than mention-edges. Total edge
load is the strongest single signal that the network already
treats a species as a hub.
Species that appear under more than one agent's
Species_<Agent> label, or that connect to the
central SharedSpecies hub via
MANIFESTS_AS, are by definition relevant beyond a
single river or forest. The network as a whole has voted.
The FEEDS in-degree (how many species this one
supports) plus a bonus when trophic_role contains
"apex", "keystone", "foundation" or "ecosystem engineer".
Predators and producers both score; passive consumers don't.
log(1 + facts_attached + publications_attached).
A species the curator returned to many times is, demonstrably,
one the curator found load-bearing — provided we read this as
attention paid, not as truth proven.
The number of EcologicalEvent nodes that stress it
plus the number of PolicyContext nodes that protect
it. Species at the centre of management decisions or active
threats score higher here.
Count of distinct HABITAT edges. Breadth without
hub-status alone is weak signal — which is why this component
has the smallest weight — but a species using many habitats is
usually doing real work in each.
Each agent's curated Species_<Agent>
set, with the top three by centrality. Click an agent to see the
full ranked list and per-species component breakdowns.