Conservation & Fieldwork·Guide

Publishing sensitive species

Coordinate generalization is standard conservation practice, not a paranoid platform feature. GBIF has maintained formal guidance on generalizing sensitive species-occurrence data since 2006 because uncontrolled distribution of exact locations for rare or commercially valuable species has documented poaching consequences. Softening a pin is still a judgment call with limits, not a guarantee that a motivated actor cannot narrow the search.

Before anything goes public

Five decisions that beat a vague “be careful”

  1. Is exact location the story?

    Donors and partners often need habitat context, not nest GPS. Prefer redacted or regional map precision when the taxon is high-target.

  2. Treat obfuscation as mitigation

    Randomizing within a radius is not anonymity. Public imagery and simple heuristics can still shrink search areas dramatically.

  3. Species panel is opt-in

    Public observations on photo detail default off. Turning the panel on is a deliberate publish act, not a side effect of uploading.

  4. IUCN as judgment, not a badge

    VU, EN, and CR should raise the bar for what notes, organism IDs, and map precision ever leave the studio, even when the species panel is on.

  5. Station coords stay internal

    Deployment / station coordinates belong to the survey workspace and partner export. They do not belong on the public photo panel.

Why generalization became practice

GBIF’s long-running guidance on sensitive primary biodiversity data exists because occurrence datasets can put animals at risk when exact coordinates travel farther than the research need. Publishing a beautiful camera-trap frame without deciding what location precision rides with it is how “open science” accidentally becomes a scout map.

ArtInStack separates the scientific record you keep for partners from what a visitor sees on a portfolio photo page or Impact Map. That split is the product expression of the same GBIF-era caution, not a novel privacy gimmick.

Obfuscation has hard limits

In a 2021 case study, Beery and Bondi showed that when camera locations were randomized within a 1 km radius, researchers using public satellite imagery and simple heuristics could shrink the likely search area by about 87% (arXiv:2106.11236). “We fuzz the coordinates” is a mitigation against casual exposure, not a guarantee against a motivated actor with imagery and time.

A related finding from Conservation Biology: among 170 peer-reviewed papers that used obscured iNaturalist occurrence data in modeling, only four acknowledged that obscuring was happening and excluded those records (Koo et al., 2025). Practitioners routinely trust a privacy mechanism without stating or understanding its actual guarantees. Your publishing protocol should not inherit that silence.

What you are deciding on ArtInStack

Public species on photo detail is a portfolio-level opt-in that defaults off, parallel to showing EXIF. When it is on, visitors can see observation identity you chose to publish; elevated IUCN categories still suppress organism IDs and internal notes on that public path. Raw deployment coordinates never appear on the public photo panel, even when species is shown.

Maps remain a separate lever: portfolio map location can stay public, redacted, or private. On plans with field surveys, **Data Governance** adds organization rules on top (automatic IUCN threat tiers, species overrides, and survey embargoes) so Evidence, Impact Map pins for conservation observations, and GBIF or Camtrap packages share one cascade above Geo privacy.

IUCN as a publishing bar

Treat Vulnerable, Endangered, and Critically Endangered as a judgment surface: raise the bar for organism identifiers, free-text notes, and map precision before anything leaves the studio. Taxonomy labels are not a substitute for that call; they are a prompt to make it.

Recommended default for elevated taxa: keep the public observations panel off until the story is ready; prefer redacted or regional pins on public maps; enable Data Governance automatic threat obfuscation when you want enforcement without a per-photo checkbox; never treat partner Camtrap export as interchangeable with the public CDN path.

Handoff to how-tos

For org rules, use Data Governance for conservation. For toggles and map settings, use Public observations on photo detail and Impact maps and location privacy. For the connective-tissue story that gets you here, start at Field conservation’s two stacks.

Mitigation, then a protocol

Softening coordinates is necessary and incomplete. Decide panel, map, and notes before elevated taxa go public, then use Data Governance and the map docs for the switches.

Also in Guides: Publishing to GBIF: endorsement first · Evidence storytelling for skeptical donors · Why open standards (Darwin Core)

All Conservation & Fieldwork guides · Guides home