Each of the sampling-event datasets published through GBIF.org now includes a complete list of the individual surveys or sampling events that it contains. A table displaying these individual ‘events’ appears below the dataset map and provides direct links to both overviews of the events and the sets of occurrence records connected to them. This addition is the first in a series of stepwise improvements to GBIF’s data indexing processes that will improve the integration and display of ecological monitoring data on GBIF.org.
The sampling-event format provides a vehicle for ecologists to share their data while adequately representing details such as sampling effort, the presence or absence of species, and attributes of abundance. Researchers who download such data can use it to produce complex analyses about broad-scale ecological processes and changes. Taken together, the event list and other upcoming enhancements will make it easier to explore, visualize and use higher-quality time-series, atlas and standardized survey datasets as groups of comparable sampling events.
Sampling-event datasets come from thousands of different kinds of environmental, ecological, and natural resource investigations. The ‘events’ themselves can include anything from occasional field surveys to systematic long-term monitoring and assessment efforts—freshwater and marine sampling, plant cover and vegetation plots, and citizen science bird and butterfly counts, among others.
Example sampling-event datasets
- Demersal and pelagic species of fish and squid from the Patagonian shelf, published by ArOBIS, the Argentinean node of OBIS
- PINK - Dragonfly (Odonata) monitoring for the permanent surveillance of coastal areas in Flanders, Belgium, published by the Belgian Research Institute for Nature and Forest (INBO)
- Freshwater samples in MZNA-INV-FRW: Macroinvertebrate samples from the water quality monitoring network along the Ebro Basin, published by the Department of Environmental Biology, University of Navarra
- Environmental monitoring of ants (Hymenoptera: Formicidae) in the influence areas of Santo Antônio Hydroelectric Power-Plant in Rondônia, Brazil, published by SiBBr: Sistema de Informação sobre a Biodiversidade Brasileira
In such investigations, researchers use consistent methods and approaches—systematic arrangements of sampling sites, fixed intervals of time, standardized traps—to collect richer, more quantitative data that may include species abundance and frequency of occurrence.
Sampling-event data supports more complex statistical analyses than can be achieved using ‘presence-only’ evidence (see below). Repeated counts or measurements from well-planned surveys, and even the absence of a particular species from a given sample, make it easier for data users to compare data from different times and places. In particular, sampling-event data can support estimates of population size, abundance and, especially population dynamics, offering the possibility of not only describing patterns of biodiversity, but also exploring ecological processes and global change.
Learn more: Introduction to sampling-event data
GBIF, its partners in the EU BON project and other stakeholders developed the Event Core extension to the Darwin Core standard to enable sharing of such data from and with ecological researchers around the world. Since support for sampling-event datasets was introduced in 2015, a growing number of data publishers have adopted the format for sharing their data.
Key GBIF partner networks have also adopted the Event Core data standard. The Ocean Biogeographic Information System—better known as OBIS—promotes it in conjunction the enhanced Measurements and Facts extension they have developed. Meanwhile, GEO BON is promoting use of sampling-event datasets to mobilize the high-quality survey data needed to support of the Essential Biodiversity Variables EBVs.
Why does this step matter?
Each of the more than 1 billion occurrence records published through the GBIF network offers evidence of the presence of a specific organism at a particular time and place. The relative ease of publishing ‘presence-only’ data has enabled a collaboration between institutions and individuals sharing data of an unprecedented scale and value to science and society. Presenting this evidence in a standardized way, regardless of its source or how and where it was gathered, has allowed researchers to advance our understanding of species distributions as well as the roles that environment, climate, land use and many other factors play in structuring biodiversity patterns.
Many of the datasets that contribute to GBIF derive from such survey activity, but until relatively recently GBIF did not offer researchers the tools to share the necessary elements to represent individual sampling events, with references to the methods and protocols used, or to document the relative abundance of each species in these samples or, where relevant, the absence of a species of interest from a sample.
At present, individual event pages do not include information on species that were absent from a given sampling event, and they may not yet show the event’s full date range. In coming months, GBIF will address these restrictions and will also show counts or measures of abundance instead of the number of occurrence records for each species. Once the full suite of enhancements is in place, we expect to prepare additional overviews and visualizations for the sampling events within each dataset and to build tools for exploring events from different datasets that used the same survey protocols to study the same taxonomic groups.
GBIF welcomes feedback from data users and publishers about these changes as well as recommendations on how best to improve access to sampling-event data within GBIF.org. We encourage all relevant research communities to adopt this approach and to make their data as openly accessible as they can for the broader scientific community.
Please contact helpdesk@gbif.org if you need help or wish to learn more about sharing sampling-event data.