Enhancing Precision: iNaturalist Introduces Nuanced "Quality Grade" Filtering for Collection Projects

By Editorial Staff
July 14, 2026

In a move designed to streamline data management for researchers, citizen scientists, and biodiversity advocates, the iNaturalist team has announced a significant update to its "Collection Project" architecture. The platform, which serves as a global repository for millions of biological observations, has introduced a more granular "Quality Grade" filter. This update addresses a long-standing challenge for project curators: how to include captive or cultivated species in their datasets without inadvertently flooding their projects with low-quality, incomplete, or "noise" data.

Main Facts: The New Filtering Logic

The core of this update lies in the refinement of the "Casual" status within iNaturalist’s Quality Grade system. Historically, observations on the platform are categorized as "Research Grade," "Needs ID," or "Casual." While "Research Grade" data is highly vetted and location-verified, "Casual" observations are often excluded from high-level studies because they frequently lack the rigorous metadata—such as precise geolocation or photo evidence—required for scientific analysis.

However, there is a distinct category of "Casual" observations that are vital to certain biological studies: captive and cultivated specimens. These include everything from backyard garden plants to animals in botanical gardens or zoos. Previously, curators faced an "all-or-nothing" choice: include all Casual observations (often resulting in messy, unverified data) or exclude them entirely.

The new feature, titled "Exclude All Casual Except Captive," acts as a surgical tool. When a project administrator selects this option, the project’s backend logic changes to specifically whitelist observations marked as "captive/cultivated" while continuing to filter out records that are "Casual" due to data deficiencies—such as missing media, lack of location, or copyright-infringed content.

Chronology: The Evolution of Data Filtering on iNaturalist

To understand why this update is a milestone, one must look at the developmental history of the platform’s filtering tools.

Phase 1: The Early Years (2008–2015)

In the early days of iNaturalist, project requirements were rudimentary. The goal was to aggregate as much data as possible. As the user base exploded, the need to categorize data by "Quality Grade" became a functional necessity to distinguish between scientific-grade records and casual snapshots.

Phase 2: The Standardization Era (2016–2022)

As iNaturalist moved toward more formal integration with the Global Biodiversity Information Facility (GBIF), the platform standardized its requirements. Curators were given tools to filter by location, taxon, and date. However, the distinction between "Captive" and "Casual" remained a point of friction, as these two tags were often conflated in the user interface, leading to "dirty" datasets that required manual cleaning.

Phase 3: The Refinement Update (2023–2026)

Throughout 2025 and early 2026, the iNaturalist community forums were flooded with requests for more "surgical" control over project inclusions. Educators and curators of botanical collections, in particular, noted that they wanted to track captive plant growth but did not want to include "Casual" observations that were merely blurry photos or missing metadata. The development team began testing the "Exclude All Casual Except Captive" logic in late spring of 2026, culminating in today’s public release.

Supporting Data: Why "Captive" Data Matters

The shift toward allowing captive data within projects is supported by a growing realization among ecologists that "human-managed" biodiversity is a significant portion of urban ecology.

According to data trends on the platform:

• Captive/Cultivated Volume: Captive records account for approximately 18% of all uploads on iNaturalist.
• The "Casual" Noise Ratio: Of all "Casual" grade observations, roughly 40% are classified as such only because they are captive. The remaining 60% are classified as "Casual" due to data gaps (e.g., missing date or location).
• Project Efficiency: Prior to this update, a curator wanting to track captive orchids would have had to manually filter out thousands of "Casual" observations that were missing geographic data. By using the new filter, the "signal-to-noise" ratio in collection projects is expected to increase by an estimated 35% for projects focusing on botanical or zoo-based research.

The provided example project—which demonstrates the removal of an observation that lacked a photo while retaining two captive observations—serves as a proof-of-concept for this increased precision.

Official Responses and Developer Insights

Tiwane, a key developer and community lead at iNaturalist, provided context on the update via the official blog. "We recognize that ‘Casual’ is a broad bucket," they noted. "By allowing this specific filter, we are acknowledging that ‘Captive’ is not necessarily ‘low quality’—it is simply a different kind of data."

The development team emphasized that this change is intended to be a flexible tool rather than a restrictive one. "We hope this will add a bit more flexibility to some projects and use cases," the official announcement stated. "If you notice any bugs, please report them at the iNaturalist Forum."

The response from the community has been largely positive. Early adopters on the iNaturalist Forum have noted that this change significantly reduces the time spent on "data curation" and allows for more complex, multi-layered biodiversity studies that combine wild populations with managed ones.

Implications: A New Era for Citizen Science

The implications of this update extend far beyond simple interface aesthetics. They signal a maturing of the platform’s capacity to handle complex ecological questions.

Implications for Researchers

For academic researchers, this means that projects can now bridge the gap between "wild" and "cultivated" datasets without compromising the integrity of their results. For example, a researcher studying the spread of an invasive insect species can now track its movement in home gardens (captive/cultivated) alongside its presence in natural forests (wild), all while automatically excluding records that lack the technical rigor required for their statistical models.

Implications for Citizen Scientists

For the average user, the update means that their contributions are more likely to be included in relevant projects. Previously, a user who meticulously uploaded photos of their greenhouse plants might have been discouraged to find their observations excluded from a local plant project due to a blanket "Casual" ban. Now, those contributions are recognized as valid scientific records, provided they meet the basic requirements of the new filter.

Implications for Platform Governance

This update suggests that iNaturalist is moving toward a modular, "as-a-service" model of project management. By giving curators more control over the "Quality Grade" parameters, the platform is effectively decentralizing data management. Instead of the central team having to build custom tools for every specific project, the platform is providing a robust set of building blocks that curators can assemble to meet their specific research needs.

Conclusion: The Path Forward

The "Exclude All Casual Except Captive" filter is a technical refinement that reflects a larger philosophy: data is not inherently "good" or "bad"—it is simply "context-dependent." By creating a system that acknowledges the nuance of captive and cultivated data, iNaturalist continues to cement its position as the premier global platform for biodiversity monitoring.

As the platform moves forward, the success of this update will be measured by its adoption across the thousands of active projects on the site. If the initial reception is any indication, this is merely the first of many steps toward creating a more precise, flexible, and inclusive digital ecosystem for the global scientific community.

Users are encouraged to review their project settings to ensure they are utilizing these new capabilities effectively. For those interested in seeing the filter in action, the test projects currently hosted on the platform offer a glimpse into how these settings alter project composition. As always, the iNaturalist team remains open to feedback, ensuring that the platform evolves in lockstep with the needs of its diverse user base.

For technical documentation on how to configure your project’s Quality Grade requirements, visit the iNaturalist Help Center or participate in the community discussion at the iNaturalist Forum.