Automating the use of citizen scientists’ biodiversity surveys in iNaturalist to facilitate early detection of species’ responses to climate change

A BioBlitz is a field survey method for finding and documenting as many species as possible in a specific area over a short period. The National Park Service and National Geographic Society hosted the largest BioBlitz survey ever in 2016; people in more than 120 national parks used the iNaturalist app on mobile devices to document organisms they observed. Resulting records have Global Positioning System (GPS) coordinates, include biological accuracy assessments, and provide an unprecedented snapshot of biodiversity nationwide. Additional processing and analysis would make these data available to inform conservation and management decisions. This project developed a process to integrate iNaturalist data with existing species lists from national parks (fig. 2). This cross-referencing of citizen science data with species inventories helps to update taxonomic and species occurrence information and may detect species range shifts or new invasions. Further, this process can serve as the basis for incorporating other online databases and to increase engagement of the public in biodiversity stewardship.



Principal Investigator : Erin E Boydston

Cooperator/Partner : Vijay Barve, Lena Lee

Co-Investigator : Toni Lyn Morelli, Jenny S Briggs



Project Report

1. Developed and tested new computer code, named TaxaTaxi

• The TaxaTaxi code compares taxonomic observations from online biodiversity databases to information about known biodiversity for a particular area. The code outputs information on potential taxonomic updates, proportions of biodiversity observed, and potential new species detections for national parks or other defined areas that may indicate species range expansion. For this CDI-awarded project, we focused on iNaturalist and NPSpecies databases.
• To test and refine our code, we compared TaxaTaxi output to results for 9 national parks for which we ‘manually’ compared NPSpecies and iNaturalist through close inspection of records. We also used the taxonomic expertise within our collaboration to inspect outputs from TaxaTaxi for possible further refinements.
• We encountered several challenges inherent in working with species information from so many different park units. One of the solutions was to build in more flexibility to our TaxaTaxi code, to allow end-users to filter and query based on their particular information needs.
• That online biodiversity databases are updated and grow through a wide community of users is one of the reasons to access and utilize such publicly-available information. Challenges from using online data though include that internal rules for accuracy assessments or other definitions may change, or there can be greater restructuring that could prevent automated code functions from accessing the data.
• Next steps in the development of TaxaTaxi include:
  1. Continued testing and refinement of the code, and potential migration to R
  2. Further discussion with (1) NPS partners about NPSpecies and BioBlitzes to verify TaxaTaxi outputs and understand park-specific inputs (2) developers and data managers for iNaturalist and ITIS (3) developers of the R-package Taxize for how to best incorporate Taxize into TaxaTaxi (4) USFWS developers of similar code
  3. Building a front-end, user friendly tool for determining inputs and filtering / querying outputs.

2. Coordination with DOI partners

• Met with existing park partners for feedback on TaxaTaxi and information about park resources
  
  Developed new NPS partners at local and national levels, including NPSpecies managers
  
  Communicated with USFWS about code recently developed for National Wildlife Refuges species information that incorporates online public databases such as iNaturalist.
  
  Wrote USGS-NPS ServiceFirst grant stemming from this project “Using Citizen Science to Enhance Natural Resource Stewardship: a Collaboration between NPS and USGS” (not awarded)

3. Obtained and explored National Park Service visitor usage data, to relate to BioBlitz efforts and findings.

4. Presentations given

• Boydston, Briggs, Morelli, Lee. 2017. Public presentation. Preliminary Biodiversity Results of the Centennial BioBlitz, including Santa Monica Mountains National Recreation Area. Santa Monica Mountains NPS Science Day, Thousand Oaks, CA.
• Morelli, Boydston, Briggs, Lee. 2017. Contributed oral presentation. One Hundred Parks and Counting: Biodiversity Findings and Outreach Impacts from a Nationwide BioBlitz. George Wright Society Conference, Norfolk, VA.
• Boydston, Backiel, Barve, Blackburn, Briggs, Morelli. Poster. 2017. Using input from citizen scientists to monitor biodiversity and detect species range shifts. CDI Annual Meeting, Federal Center, Denver, CO.

5. Began draft of a manuscript, draft title “Public involvement in biodiversity stewardship of national parks”



6. Related links

• NPSpecies: https://irma.nps.gov/NPSpecies/
• iNaturalist 2016 BioBlitz data: https://www.inaturalist.org/projects/2016-national-parks-bioblitz-nps-s…

Figure caption for attached figure:

Graphic depicting TaxaTaxi Computer Code: TaxaTaxi compares online biodiversity databases to find species in common and unique to each database. Here, the inputs for comparison were (A) species lists for National Parks, called NPSpecies, and (B) observations of organisms from the 2016 National Park Service-National Geographic BioBlitz that were recorded in iNaturalist. The first filter (white oval with gray dots at top) checked if species in iNaturalist were present in NPSpecies for the park where the observations in iNaturalist were made. Records output by this filter as unique to iNaturalist (red container on right) were checked for taxonomic synonyms through comparison to ITIS and incorporation of R-package Taxize. Because scientific names often change, identifying taxonomic updates can assist curators of biodiversity information. After aligning taxonomies between lists, filtering the inputs a second time narrows the output to species not present in NPSpecies (orange container on right). These may be records of species not known to occur and could signal important ecological changes. TaxaTaxi also identifies species in NPSpecies that were not recorded in the 2016 BioBlitz (green container, left). These results can indicate modifications to consider for future citizen science efforts. Species in common (gray container, right) verify organisms from NPSpecies present in the 2016 BioBlitz and show the proportion of a park’s biodiversity recorded through citizen science effort.