📊 Sweden — GBIF Gap Finder

This tab turns what the other tabs found into a clear, prioritised to-do list — where records are missing, and what to do about it. It is built for the people who decide where time and money go: GBIF node staff, collection managers, and data coordinators.

The Recommended Actions below are concrete, countable goals, grouped by the kind of gap each one closes:

• Spatial: Grid cells with no GBIF records are the highest priority. Stale cells (no GBIF records newer than 5 years) may warrant resurvey — after checking whether recent data exists outside GBIF.
• Taxonomic: Missing threatened species (CR/EN with zero GBIF records) are critical for conservation. Under-sampled orders (high species richness, low occurrence count) indicate systematic collection biases.
• Species of Concern: Sensitive species with degraded coordinates, invasive species lacking monitoring data, and threatened species without any GBIF records all require targeted attention.
• Temporal: Cells and taxa with only historical records and no recent activity may represent abandoned monitoring programmes or shifted survey effort.
• Infrastructure: Single-publisher cells and under-diversified geographic regions indicate data resilience risks. Use the Publishers tab to identify taxonomic groups where additional data sources are needed.

The Next 12 Months section projects realistic targets based on recent performance. Targets are set at 1.5× the rate achieved in the last 12 months — ambitious but achievable. These projections help frame discussions with funders, data holders, and institutional partners about what is possible with sustained effort.

Use the Export button to download the priority lists as a spreadsheet for sharing with stakeholders, incorporating into grant proposals, or feeding into institutional work plans.

This tab shows how biodiversity observations are distributed across the country's 10 km EEA reference grid cells. Each cell is coloured by the selected metric: total occurrences, data recency (how recently each cell was surveyed), species richness, or observations from the last 12 months.

Use the Kingdom filter to isolate specific taxonomic groups. Bird observations dominate Swedish GBIF data, so filtering to non-Aves groups can reveal sampling gaps that are otherwise hidden. The Class filter allows further refinement within a kingdom.

Data recency shows how stale each cell's most recent observation is. The palest cells have no GBIF-mediated records dated within the last 10 years — recent data may exist outside GBIF, so check national/regional sources before prioritising resurvey. Light cells (5–10 years) are approaching staleness. The darkest cells have data from the last year; grey cells have no data at all.

Occurrence distribution (histogram) shows how records are spread across cells. A healthy dataset has a smooth distribution; a spike at the low end indicates many cells with only token data (1–10 records), which may be insufficient for ecological analysis.

Grid cells are based on the European Environment Agency (EEA) reference grid at 10 km resolution. The grid is clipped to the country boundary using GADM administrative boundaries.

This tab shows when biodiversity observations were made. The historical trend shows total occurrences per year; the seasonal pattern reveals monthly collection biases.

The heatmap shows year × month intensity. Switch between log scale (better for spotting patterns across orders of magnitude) and linear scale (better for comparing absolute numbers). Use the taxonomic filters above to isolate specific groups.

The sharp increase in recent decades is largely driven by citizen science (especially Artportalen/iNaturalist). Filtering by kingdom or order can reveal which groups are driving temporal trends.

What this tab measures: GBIF-mediated occurrence data assessed against the national taxonomy backbone (Dyntaxa). Unlike the Spatial, Temporal, Record Types and Publisher tabs — which show all GBIF records for Sweden with no reference filter — every completeness and gap figure here is relative to the national checklist: of the species Dyntaxa lists, how many have GBIF records, and how sampling effort is distributed across groups.

The Taxonomic Bias chart (following Troudet et al., 2017) reveals whether groups are over- or under-represented relative to their known species richness. If a group has 10% of all known species but only 1% of all occurrences, it is under-sampled. The default landing view uses GBIF-style groups — curated mixed-rank categories (Birds, Mammals, Insects, Vascular Plants, Fungi, etc.) that match how GBIF's country pages present data. Switch to 'Kingdoms' for the standard taxonomic hierarchy.

The Exclusion filter lets you remove dominant groups (e.g. Aves) from the bias chart to reveal patterns among less-sampled taxa. The cascade filters (Kingdom → Phylum → Class → Order → Family) let you drill into any group, and the active filter breadcrumb shows your current drill-down path.

The Last 12 Months toggle highlights recent observed sampling effort, showing whether recent data collection is addressing historical biases or reinforcing them. When toggled on, the species count chart displays the number of occurrences observed in the last 12 months as annotations to the right of each bar. For sub-population views (native, introduced, invasive, threatened species), see the Species of Concern tab.

Species Count by Order and Species Coverage by Family show how many species in each group are present in GBIF vs the national backbone. Green bars indicate species found in GBIF; sand-coloured bars show missing species.

Reference population: completeness percentages use the species-rank taxa in the national checklist (Dyntaxa) as the denominator — excluding microbial kingdoms (Bacteria, Archaea, Viruses) and Homo sapiens. A species counts as “in GBIF” when at least one occurrence resolves to it.

This tab focuses on three categories of species that require special attention for conservation monitoring and data mobilisation:

• Threatened — species on the national Red List, categorised by the IUCN framework as Critically Endangered (CR), Endangered (EN), Vulnerable (VU), Near Threatened (NT), or Data Deficient (DD). A 'missing' threatened species is one that appears on the Red List but has zero matching GBIF occurrence records. These are the highest conservation data priority: without occurrence data, range modelling, population trend analysis, and habitat suitability assessments cannot be performed.
• Invasive — species listed as invasive in GRIIS, the Global Register of Introduced and Invasive Species for Sweden (only taxa GRIIS explicitly flags as invasive, not all introduced or alien species). Because GBIF occurrence cubes are aggregated at the species level, this flag is applied at the species level too: a species is marked invasive if any of its listed forms — including a subspecies or variety — appears as invasive in GRIIS. The count therefore answers “which species have an invasive form that may warrant monitoring” , not whether one specific subspecies is invasive. Monitoring invasive species requires spatially and temporally complete occurrence data to detect range expansion, evaluate management interventions, and trigger early-warning alerts. Species missing from GBIF represent blind spots in the national invasive-species surveillance network.
• Sensitive — species whose precise location data is restricted in GBIF to protect them from collection pressure, habitat disturbance, or trade. These records may show generalised coordinates (e.g. country centroid or 50 km grid) rather than exact localities. This affects the accuracy of spatial gap analysis for these species, and the true distribution may be much better known than GBIF data suggests.

The taxonomy cascade filters at the top apply across all three sub-tabs. Use the Scope filter to restrict to native, introduced, or invasive species (where establishment-means data is available). Threat status comes from SLU Artdatabanken's Red List, the invasive flag from GRIIS Sweden, and the sensitive flag from the SLU Restricted Access Species list. The Red List and sensitive flags are matched at the rank each list publishes — a listed subspecies stays a subspecies — whereas the invasive flag is rolled up to species level, as described above. A few non-species entries (hybrids, colour morphs, slash-aggregates) have no species-level occurrence equivalent and are not flagged.

Prioritisation tip: start with the Threatened sub-tab to identify missing CR/EN species, then check the Invasive sub-tab for unmonitored invasive species in the same taxonomic groups. Species that are both threatened and invasive (e.g. a threatened native species in a genus with invasive congeners) may warrant especially urgent data mobilisation.

Reference lists in use — resolved to the exact GBIF-published datasets; the title below confirms each edition, and the DOIs travel in the data bundle:

This tab shows which organisations publish biodiversity occurrence data to GBIF for this country. Understanding the publisher landscape helps assess data infrastructure resilience, identify potential data partnerships, and recognise under-represented data holders.

Taxonomic filter: Use the kingdom/class/order filters to see which publishers contribute data for specific taxonomic groups. This reveals whether bird data comes mainly from citizen science while insect data depends on museum collections, for example. The dependency map also updates to show per-cell publisher coverage for the selected group.

Publisher category: Publishers are classified by name into three categories: Citizen science (Artdatabanken/Artportalen, iNaturalist, eBird, etc.), Private sector (environmental consultancies and companies), and Research data (universities, museums, herbaria, government agencies, sequencing facilities, field stations and marine institutes). Bars in the charts are colour-coded by category. Choosing a category also limits the dependency map to publishers of that type.

Publisher Dependency per Cell maps each 10 km grid cell by the number of distinct publishers contributing data. Cells with a single publisher are both an infrastructure vulnerability and a partnership opportunity — if that organisation paused contributing, the cell would lose coverage, so broadening the contributor base safeguards it. This reflects the publishing infrastructure, not the publisher.

The All Publishers table shows every contributing organisation with their occurrence count, species count, category, and percentage share.

Each GBIF occurrence record has a basis of record describing how the observation was made. The main types are: Human Observation (field sightings, citizen science), Preserved Specimen (museum/herbarium collections), Machine Observation (camera traps, acoustic sensors), Fossil Specimen (paleontological collections), and Material Sample (DNA, tissue, environmental samples).

Occurrences by Basis of Record shows the overall share of each record type. Use the 'Last 12 Months' toggle to see how recent data collection compares to historical records. When toggled on, bars show prior records (faded) and recent additions (solid).

Temporal Trend shows the time series for a single selected basis type. The sharp increase in Human Observations since ~2010 reflects the growth of citizen science (primarily Artportalen and iNaturalist).

Spatial Coverage shows what percentage of Sweden's 10 km grid cells have at least one record of each type. Human Observations cover the most cells; museum specimens are concentrated in fewer areas.

Species Coverage shows the total number of species detections summed across all cells. Note: a species recorded in 3 cells counts as 3, not 1. This measures sampling breadth, not unique species count.

Spatial Distribution maps where each basis type has data, using binned occurrence categories.