Belgian-built GBIF Alert system wins the 2023 Ebbe Nielsen Challenge

Species occurrences notification system leads the way, with Frictionless Data improvements to Bionomia earning second place and third place shared by Library Identification Resources and Open Data Biodiversity Mapper

Ebbe-2023
Illustration of a Japanese horseshoe crab Tachypleus tridentatus(Leach, 1819) from Fauna japonica v.1 Crustacea (1823-1830) via Biodiversity Heritage Library, no rights reserved under CC0.

GBIF Alert, a flexible open-source system for notifying users about the availability of newly occurrence records through GBIF, has earned selection as the first-prize winner of the 2023 GBIF Ebbe Nielsen Challenge.

The members of the Belgian team responsible for GBIF Alert, led by Nicolas Noé of The Binary Forest, are no strangers to the GBIF community or the Challenge winners' podium. Noé was part of a three-person team that won the Challenge's inaugural prize in 2015 along with Peter Desmet from the Research Institute for Nature and Forest (INBO), who has joined him again here. Lien Reyserhove and Damiano Oldoni of INBO likewise partnered with Desmet on their 2018 joint first-place entry. INBO colleagues Tim Adriaens and Bram D’Hondt round out the members of this year's winning team.

Frictionless Data from Bionomia, submitted by David Shorthouse, won second prize for a new export feature that supports ongoing improvements to data quality for specimen records from scientific collections. Two entries that will share third place round out the list of winners for 2023:

The jury, led by Birgit Gemeinholzer, professor of botany at the University of Kassel and current chair of the GBIF Science Committee, selected this year's Challenge winners from a group of 12 qualified submissions. This annual incentive prize honours the memory of Dr Ebbe Schmidt Nielsen, a Danish-Australian entomologist who was one of the principal founders of GBIF and an inspiring leader in the biosystematics and biodiversity informatics community.

Allocations from the annual prize pool of €20,000 will award the first-place team with €8,000, €6,000 for second place, and €3,000 each for the third-place winners.

2023 GBIF Ebbe Nielsen Challenge winners

First Prize
GBIF Alert: an open-source GBIF-based alert system for occurrences

This reusable open source tool demonstrates how to use GBIF.org as a system for notifying users of newly available occurrence records for any species or location of interest. While it can be deployed and configured to generate alerts for any combination of specific taxa, areas or source datasets, the system's most likely use case focuses on the management of invasive alien species (IAS).

Field managers and policymakers responsible for monitoring the spread of IAS require rapid access to information about the appearance of new occurrences of species of concern, whether in known areas of introduction or new ones. GBIF Alert can keep them informed about new (or new to GBIF) records of their species of interest, triggering email notifications selected through a simple, intuitive user interface. The demo submitted for the Challenge is configured to monitor the GBIF occurrence index for new records of invasive species in North America.

"GBIF Alert arose directly from the experience of developing Life RIPARIAS, an open-source GBIF-based early alert system targeting some invasive riparian plants and crayfishes in parts in Belgium. Once we realized that a more flexible, less hard-coded configuration would make it useful to many other teams around the world, it just felt wrong to keep it to ourselves, so we put in some extra effort.” said Noé. “Our happiness and surprise at winning first prize has given us a boost that we'll invest in other biodiversity informatics tools in the same open-source spirit, and we'd love to see the functionality of GBIF Alert integrated directly into GBIF.org."

video | GitHub | GBIF Alert demo for U.S. invasives

Second Prize
Frictionless Data from Bionomia

This entry prepared by Canadian biodiversity informatician David Shorthouse stands alone as second-prize winner of the 2023 Challenge. Bionomia, the platform he first released five years ago, leverages persistent identifiers from ORCID and Wikidata to enable volunteer "scribes" to annotate specimen records shared into GBIF.org, clarifying the credits they give to individual specimen collectors and identifiers. In the process, it has sparked wider discussions about better recognizing and rewarding the expertise required to assemble, maintain, digitize and maintain the knowledge from the world's scientific collections.

Shorthouse's 2023 Challenge entry has outlined a new feature that automatically prepares Frictionless Data Packages of collector and identifier attributions in Bionomia for every dataset. Relying on this simple, convenient format makes it easier for a collections manager to import Bionomia attributions into a local collection management system and share newly updated and improved records, creating a virtuous "round trip" cycle of data quality improvements.

The jury's selection of Bionomia highlights the value of building new functionality into existing tools. Meanwhile, the data managers who are currently improving the data quality of their collections with Bionomia attributions will share pointers and lessons learned on 6 December 2023 during a half-day virtual workshop on round tripping, for which registration is now open.

"These new outputs help demonstrate the institutional value of Bionomia and offer a positive return on investment for specimen digitization and transcription," said Shorthouse. "Persistent unique identifiers for people may seem lightweight items to inherit, but they serve as passports into the provenance, microhistories and emotive threads that link the people who laid the foundations of our natural history collections with their support networks, what they collected, where they were, and our shared planet. I hope all collections that share data into GBIF can use Bionomia's round-trip service to help strengthen their own local networks."

video | GitHub

Third Prize
Library of Identification Resources

This web-based platform developed by biologist and programmer Lars Willighagen combines two web-based services that guide citizen scientists toward suitable taxonomic keys to help them identify the organisms that they themselves observe. Willighagen is currently enrolled as a Master's student at the Radboud Institute for Biological and Environmental Sciences (RIBES) of Radboud University in Nijmegen, the Netherlands.

A machine-readable catalog of public-domain taxonomic reference works ranging from simple pictorial guides to comprehensive serial publications provides the first underlying service. This bibliographic repository contains metadata about the reference works and information about their taxonomic and geographic scope and completeness. Where possible, the full list of names appearing in the works are extracted and mapped to the GBIF Backbone Taxonomy using the GBIF API.

Users tap into the more visible second service by entering the location and scientific name of an observation in need of identification. The search results returns potentially relevant reference publications in the catalog and estimates of how useful they may be in verifying the identity of the observation. Suitability estimates rely on the GBIF API to compose a checklist of all species found at the geographic search location, which can be compared to the publications' taxon lists. Use of the available data accommodates the changes in species distribution over time, given that older keys focused on specific taxa from target regions may become unsuitable while others restricted to different regions have become suitable.

video | Find Identification Resources | Catalog | GitHub: Resource Search | GitHub: Catalog

Third Prize
Open Data Biodiversity Mapper

The R-based Shiny webapp prototype built by Sam Perrin, Philip Stanley Mostert and Ron Togunov of NTNU runs an open-source data pipeline that preprocesses different types of datasets, making it easier for users to produce integrated species distribution models (ISDMs).

This approach recognizes that, despite the benefits of standardization, opportunistic observations by citizen scientists, for example, differ significantly from records collected using a complex protocol for sampling abundance. By preserving some of the unique strengths and characteristics of the constrasting collection methods in underlying datasets and accommodating them within a single statistical framework—rather than simply pooling records and flattening their differences—Open Data Biodiversity Mapper improves resulting ISDM visualizations.

The tool currently generates species intensity maps for four different taxonomic groups in Trøndelag county in central Norway, but the team is currently scaling the tool up to produce maps for all of Norway. While intended to make the ISDM approach accessible to groups without the necessary data wrangling skills, access to the scripts and step-by-step instructions for using them should enable users with some knowledge of GitHub, GBIF and R to create localized versions of the tool for other target taxa.

video | GitHub

Jury for 2023 Ebbe Nielsen Challenge