Species Threat Explorer

Data Overview

Top 5 Threat Categories

Top 5 Species by Threats

Species Threat Profile

Select Species for Analysis

Triplet Review & Validation

Help improve data quality by reviewing randomly selected triplets for accuracy and relevance.

Start Review Session

Review Progress

Reviews Completed: 0
Export All Reviews (CSV)

Systemic Risk Analysis

Ecological Network Visualization

Systemic Risk Metrics

Network Modularity

-
Community structure strength Ratio of edges within communities to total edges.

Network Nestedness

-
Specialist-generalist pattern Measures if specialist threats are subsets of generalist threats

Ecosystem Vulnerability

-
Overall systemic risk score Composite of avg. threats, density, and vulnerability variation

Critical Species Count

-
High-impact species nodes Number of species in the top 10% of network connections

Vulnerability Distribution

Threat Propagation Network

Critical Nodes Analysis

Vulnerability Corridors

Threat Embeddings Analysis

Explore the semantic space of threat descriptions using dimensionality reduction, clustering, and interactive visualization.

8

Dataset Information

Load embeddings data to see dataset statistics.

Interactive Threat Landscape

Select a dimensionality reduction method and click "Load Embeddings" to begin visualization.

Click on points in the visualization to see threat details.

Semantic Cluster Analysis

Perform clustering to see analysis results.

Generate cluster labels to see semantic interpretations.

Pattern Discovery

Identified Patterns

Cluster Outliers

Points that are isolated from major clusters, representing unique or rare threat descriptions.

Threat Gradients

Continuous structures indicating gradual shifts in semantic meaning across threat types.

Cluster Relationships

Spatial arrangement of clusters revealing higher-order relationships between threat categories.

Knowledge Transfer

Discover potential threats for understudied species by analyzing similar species with more complete threat profiles.

How It Works:

Initial Selection: Species are considered similar if they share a large number of threats with the target species. A combined similarity score (based on shared threat overlap) is used to rank potential source species.

Transferability Score: For each potential new threat from a similar species, a score (0-1) is calculated based on the strength of evidence (number of studies), the type of threat (broad vs. specific), and the similarity of ecological context if it's mentioned in the abstract. Scores also get additional points if the DOI shows it has been published in past 5 years to target current threats.

Target Species Analysis

0.05

Ready to Analyze

Enter a target species name above to discover potential threats based on similar species with more complete threat profiles.

Example Analyses:

Sparrowhawk

Raptor with 20% similarity to Peregrine Falcon

Barn Owl

Owl species with multiple similar candidates

Greylag Goose

Waterfowl with transferable threat knowledge