ZIP-602: Citizen Science Contribution Protocol

Abstract

This ZIP defines a protocol for citizen scientists to contribute biodiversity observations and research data to the Zoo ecosystem. Contributions are cryptographically attested at the device level, validated through a multi-tier verification pipeline, and rewarded with ZOO tokens proportional to data quality and scientific value. The protocol implements sybil-resistant reputation using decentralized identifiers (DIDs) bound to verifiable credentials, preventing reward farming while maintaining contributor privacy.

Motivation

Citizen science generates enormous volumes of biodiversity data. Programs like eBird (Cornell Lab) collect over 200 million observations annually from 800,000+ contributors. However, existing platforms suffer from limitations that blockchain-native protocols can address:

No contributor ownership: Observations submitted to centralized platforms become the property of the hosting institution. Contributors have no control over how their data is used or monetized.
No quality incentives: Beyond gamification badges, contributors receive no material reward for high-quality observations. There is no economic signal distinguishing a carefully documented rare species sighting from a casual backyard bird count.
Sybil vulnerability: Existing reputation systems (e.g., eBird reviewer status) rely on social trust and manual curation, which does not scale. Bad actors can create multiple accounts to inflate observation counts.
No interoperability: Data contributed to iNaturalist cannot be seamlessly used in eBird, and vice versa. Each platform maintains its own identity system, taxonomy, and quality standards.
Centralized moderation: A single platform operator decides what constitutes a valid observation. There is no transparent appeals process or decentralized consensus on data quality.

Zoo Labs Foundation aims to build the largest open biodiversity dataset for conservation AI. Citizen science contributions are essential to achieve geographic and taxonomic coverage that no institutional effort can match alone.

Cross-Ecosystem Context

HIP-0075 (Hanzo OSS Contributor Payouts): Token-based compensation model for open-source contributions, adapted here for scientific data contributions.
LP-3093 (Lux Decentralized Identity): DID infrastructure for sybil-resistant identity without requiring personal information disclosure.
ZIP-540 (Research Ethics & Data Governance): Ethical framework governing how citizen science data is collected, stored, and used.

Specification

1. Contributor Identity

1.1 DID Registration

Each contributor registers a Decentralized Identifier (DID) conforming to the W3C DID specification:

{
  "id": "did:zoo:contributor:<address>",
  "verificationMethod": [
    {
      "id": "did:zoo:contributor:<address>#keys-1",
      "type": "EcdsaSecp256k1VerificationKey2019",
      "controller": "did:zoo:contributor:<address>",
      "publicKeyMultibase": "z..."
    }
  ],
  "authentication": ["did:zoo:contributor:<address>#keys-1"],
  "service": [
    {
      "id": "did:zoo:contributor:<address>#observations",
      "type": "BiodiversityContributor",
      "serviceEndpoint": "ipfs://<contributor-profile-cid>"
    }
  ]
}

1.2 Sybil Resistance Tiers

Tier	Verification	Capabilities	Reward Multiplier
Tier 0: Anonymous	Wallet address only	Submit observations (unverified queue)	0.25x
Tier 1: Pseudonymous	DID + email verification	Submit observations, participate in community verification	0.5x
Tier 2: Attested	DID + verifiable credential from recognized institution	Full contribution capabilities, review flagged records	1.0x
Tier 3: Expert	Tier 2 + domain expertise credential + 500+ verified observations	Expert verification, reviewer assignment, mentor role	1.5x

Tier advancement is non-reversible (cannot be downgraded for inactivity), but reputation score within a tier can decrease.

1.3 Verifiable Credentials

Tier 2+ contributors present verifiable credentials (VCs) issued by:

Issuer Type	Examples	Credential
Academic institution	University, research institute	Researcher affiliation
Conservation organization	WWF, IUCN, local NGOs	Conservation practitioner
Government agency	Fish & Wildlife, national parks	Professional biologist
Zoo ecosystem	Zoo Labs Foundation	Trained citizen scientist (completion of Zoo training program)
Peer attestation	3+ Tier 3 contributors	Domain expertise endorsement

Credentials are verified on-chain without revealing the issuer's identity to other contributors, using selective disclosure.

2. Observation Submission

2.1 Submission Flow

Contributor opens Zoo app (mobile/web)
    |
    v
Device captures observation:
  - GPS coordinates (from device)
  - Timestamp (from device clock + NTP verification)
  - Media (photo/audio/video)
  - Species identification (AI-assisted suggestion)
  - Additional metadata (count, behavior, habitat notes)
    |
    v
Device signs attestation package:
  - Observation data hash
  - Device fingerprint (hardware attestation where available)
  - GPS accuracy metric
  - Contributor DID signature
    |
    v
Attestation submitted to Zoo L2
    |
    v
Full observation data uploaded to IPFS
    |
    v
On-chain index updated (per ZIP-601)

2.2 Attestation Schema

{
  "attestationId": "bytes32",
  "contributorDid": "string",
  "observationHash": "bytes32",
  "device": {
    "fingerprint": "bytes32",
    "platform": "string",
    "gpsAccuracy": "uint16",
    "hasHardwareAttestation": "bool"
  },
  "temporal": {
    "deviceTimestamp": "uint64",
    "ntpOffset": "int32",
    "blockTimestamp": "uint64"
  },
  "signature": "bytes",
  "previousAttestationHash": "bytes32"
}

The previousAttestationHash creates a contributor-specific hash chain, making it possible to detect if a contributor's attestation history has been tampered with.

3. Verification Pipeline

3.1 Automated Checks

Check	Method	Failure Action
GPS plausibility	Velocity check against previous observation	Flag for review
Timestamp consistency	NTP offset within 60 seconds	Reject
Species range	Known range map lookup	Flag for review
AI species match	Zoo species ID model confidence > 0.7	Auto-verify; below 0.7 flag for community
Duplicate detection	Spatiotemporal fuzzy match	Merge or reject
Media authenticity	EXIF validation, deepfake detection	Flag for review

3.2 Community Verification

Observations that pass automated checks but fall below auto-verify thresholds enter community verification:

Observation enters community queue
    |
    v
3 independent community verifiers vote:
  - AGREE (species ID correct, data plausible)
  - DISAGREE (species ID wrong, suggest correction)
  - UNCERTAIN (insufficient evidence)
    |
    v
Consensus rules:
  - 3/3 AGREE: Verified
  - 2/3 AGREE: Verified with note
  - 2/3 DISAGREE: Sent to expert review
  - All UNCERTAIN: Sent to expert review
  - Mixed: Sent to expert review

Community verifiers earn reputation for accurate verification (aligned with eventual expert determination).

3.3 Expert Verification

Expert review is required for:

New species range records (first observation of species in a region)
Observations of IUCN Red List species (Vulnerable or above)
Genetic samples
Records flagged by community verification
Records from Tier 0 contributors

Expert verifiers are Tier 3 contributors with domain-relevant expertise. See ZIP-600 reviewer reputation system for expert qualification.

4. Reputation System

4.1 Reputation Score Components

ReputationScore = BaseScore + QualityBonus + ConsistencyBonus - Penalties

Where:
  BaseScore       = count(verified_observations) * tier_multiplier
  QualityBonus    = sum(quality_score_per_observation) / count
  ConsistencyBonus = streak_days * 0.5 (max 180 days = 90 points)
  Penalties       = sum(rejected_observations * 5) + sum(sybil_flags * 50)

4.2 Quality Score Per Observation

Factor	Points	Description
Media evidence	+3	Photo, audio, or video attached
High-confidence AI match	+2	AI confidence > 0.9
Expert-verified	+5	Confirmed by Tier 3 expert
Rare species	+5	IUCN Red List species
Under-surveyed area	+3	Region with < 100 records in database
Genetic sample	+10	DNA barcode or eDNA submitted
Habitat survey data	+5	Associated habitat survey completed

4.3 Anti-Gaming Measures

Measure	Implementation
Velocity limits	Max 50 observations per contributor per day
Geographic dispersion	Observations from identical coordinates within 1 hour are deduplicated
AI-generated media detection	Deepfake detection model on submitted images
Reputation decay	2% monthly decay for inactive contributors
Stake-weighted verification	Verifiers must stake ZOO proportional to their verification volume
Graph analysis	Social graph analysis to detect coordinated sybil rings

5. Token Rewards

5.1 Reward Calculation

Rewards are distributed from a dedicated Citizen Science Pool funded by Zoo Treasury allocation (per ZIP-603 governance vote).

ObservationReward = BaseReward * QualityMultiplier * TierMultiplier * ScarcityMultiplier

Where:
  BaseReward         = 1 ZOO (governance-adjustable)
  QualityMultiplier  = quality_score / max_quality_score (range 0.1 to 2.0)
  TierMultiplier     = {T0: 0.25, T1: 0.5, T2: 1.0, T3: 1.5}
  ScarcityMultiplier = max(1.0, 5.0 - log10(region_observation_count))

5.2 Reward Distribution

interface ICitizenScienceRewards {
    function claimReward(bytes32 observationId) external;
    function batchClaim(bytes32[] calldata observationIds) external;
    function getRewardAmount(bytes32 observationId) external view returns (uint256);
    function getContributorStats(address contributor) external view returns (
        uint256 totalRewards,
        uint256 pendingRewards,
        uint256 reputationScore,
        uint256 observationCount
    );
}

Rewards vest linearly over 30 days after verification to prevent claim-and-dump behavior. Contributors with reputation scores below 10 have a 90-day vesting period.

5.3 Verification Rewards

Community and expert verifiers also earn rewards:

Role	Reward	Condition
Community verifier	0.1 ZOO per verification	Vote aligned with consensus
Expert verifier	0.5 ZOO per verification	Expert review completed
Dispute resolver	1.0 ZOO per resolution	Successfully resolved community disagreement

6. Smart Contract Architecture

CitizenScienceHub
    |
    +-- ContributorRegistry
    |       DID registration, tier management, credential verification
    |
    +-- AttestationValidator
    |       Validates submission attestations, checks device signatures
    |
    +-- VerificationEngine
    |       Manages verification queues, consensus, escalation
    |
    +-- ReputationTracker
    |       Calculates and stores reputation scores
    |
    +-- RewardsDistributor
            Computes rewards, manages vesting, processes claims

7. Protocol Parameters

Parameter	Default	Governance-Adjustable
Base reward	1 ZOO	Yes
Daily observation limit	50	Yes
Community verifiers required	3	Yes (min 2, max 5)
Verification reward	0.1 ZOO	Yes
Reputation decay rate	2% / month	Yes
Reward vesting period	30 days	Yes
Citizen Science Pool size	Set per epoch by governance	Yes

Rationale

DID-based identity: DIDs provide sybil resistance without requiring KYC. Contributors retain privacy while proving uniqueness through verifiable credentials from trusted issuers (adapted from LP-3093).
Tiered access: Lower tiers have reduced rewards but no barriers to entry. This enables broad participation while concentrating high-value rewards on verified, expert contributions.
Device attestation: Hardware attestation (where available on modern mobile devices) provides a strong signal that observations originate from a physical device at a real location, not from a script.
Multi-stage verification: Automated, community, and expert verification layers balance throughput with accuracy. Most observations are auto-verified by AI, reducing the load on human reviewers.
Scarcity multiplier: Higher rewards for under-surveyed areas incentivize geographic coverage rather than easy observations in well-studied areas.
Vesting: Token vesting prevents hit-and-run contribution patterns and aligns contributor incentives with long-term data quality.

Security Considerations

GPS spoofing: Mobile devices can report false coordinates. The protocol uses velocity checks (impossible travel speed between observations), device hardware attestation, and cross-referencing with satellite imagery to detect spoofing. Spoofed observations that pass automated checks are caught by community/expert review.
Sybil attacks: A single actor creating multiple DIDs to farm rewards is mitigated by requiring verifiable credentials for Tier 2+ (where meaningful rewards begin). Tier 0/1 rewards are too low to be economically attractive for farming. Graph analysis detects coordinated activity.
Deepfake media: AI-generated images of species are becoming convincing. The protocol runs deepfake detection on submitted media and requires EXIF metadata consistency. Flagged media requires expert review with original device verification.
Reward manipulation: The scarcity multiplier could be gamed by submitting observations in remote areas with low verification coverage. The velocity limit, vesting period, and expert review for unusual range records mitigate this.
Privacy: Contributor location history could reveal home addresses or daily patterns. The protocol stores only observation locations, not contributor locations. Contributors can opt to fuzz their observation coordinates at a personal-data level (separate from the species-sensitivity fuzzing in ZIP-601).
Credential forgery: Verifiable credentials from institutional issuers are cryptographically signed. The protocol maintains a registry of trusted issuers. Compromised issuer keys can be revoked through governance.

References

ZIP-540: Research Ethics & Data Governance
ZIP-600: DeSci Protocol Framework
ZIP-601: Open Biodiversity Database Standard
HIP-0075: Hanzo OSS Contributor Payouts
LP-3093: Lux Decentralized Identity (DID)
W3C. "Decentralized Identifiers (DIDs) v1.0." https://www.w3.org/TR/did-core/
W3C. "Verifiable Credentials Data Model v2.0." https://www.w3.org/TR/vc-data-model-2.0/
Sullivan, B. L. et al. "eBird: A citizen-based bird observation network." Biological Conservation 142(10), 2282-2292 (2009).
Bonney, R. et al. "Citizen Science: A Developing Tool for Expanding Science Knowledge and Scientific Literacy." BioScience 59(11), 977-984 (2009).

Changelog

Version	Date	Changes
1.0	2025-01-15	Initial draft

Citizen Science Contribution Protocol