Species dataset for model-based conservation planning in Latvia: occurrences, importance for nature conservation planning and ecogeographical variable selection

Description:

This dataset provides a basis for model-based conservation planning in Latvia, from species selection and observations to ecogeographical variable selection for habitat suitability modelling of species. The dataset covers multiple taxonomic groups: Amphibia, Araneae, Bryophyta, Chiroptera, Coleoptera, Diptera, Hymenoptera, Lichen, Mollusca, Orthoptera, Papilionoidea, PoroidFungi, Reptilia, Tracheophyta.

File format:

The files are provided as spreadsheets in two formats – MS Excel (.xlsx) and comma separated values (.csv). The Excel files contain a sheet with column descriptions (README). For csv versions, the column descriptions are provided in a separate file.

File structure:

All data tables contain relational key fields:

·       “Group” – a wider taxonomic group;

·       “NR” – an assigned taxon number within each taxonomic group;

·       “ScientificName” – the species scientific name;

·       “CODE” – a unique species acronym, composed of first letters of genus and species.

For each species, the combination of these fields is unique, therefore they can be used for data filtering and cross-referencing between the data tables. The other fields in each file vary and are described in the according “REAMDE” pages.

Files:

• SpeciesList – contains three sheets: Nomenclature, SpeciesList and README.
• The Nomenclature sheet summarizes 4215 nomenclature units from 4176 species, providing different synonyms used in nature conservation legislation documents, projects and recognized information sources.
• The SpeciesList sheet shows the decision-making process for species selection for distribution modelling. It contains annotations on each species' inclusion in nature conservation legislation, national red list status, conservation status as well as inclusion in multiple specialist and indicator species lists. Each species home range radius approximation is provided (best matching to any of the following groups: '500' - up to 500 m, '1250' - up to 1250 m, '3000' - around 3000 m or more), which is essential when selecting ecogeographical variables.
• field descriptions (with references where applicable) are provided in the README sheet.
• EGVtable – contains an expert-based selection of ecogeographical variables for 474 species according to their biology, ecology, home range and geographic distribution. First 8 columns contain information on the used ecogeographical variables (Avotins et al. 2025). In the columns named according to each species acronym, “1” denotes the variable selection for species distribution modelling.
• Observations – contains 161125 expert-filtered observations of 464 species. The data is pulled from diverse sources, ranging from citizen-science platforms to Nature Conservation Agency database of the Republic of Latvia, to observations gathered in research projects and expert personal data bases. The temporal coverage for most taxonomic groups is 2016 – 2024, but for some species with limited number of observations the period is extended to 2005 – 2025. The coordinates are rounded to species-specific resolution (100, 1000 or 5000 m grid cell center, denoted in field “Resolution”) to avoid revealing exact locations of sensitive species, and are spatially filtered to only keep the last known observation per respective grid cell. Species occurrences that are restricted from disclosure according to an ordinance issued by the Nature Conservation Agency of the Republic of Latvia are omitted from the observation dataset (all Chiroptera and Cypripedium calceolus L.). For each observation the year, resolution, data source, its relative reliability and a unique source database identifier, as well as annotations for possible misslocation (see FilteringGuide description) are provided. The information of the date and time of the observation, as well as individual traits (sex, life stage, status of the observation, abundance and units) are provided when available. The accumulated change of land cover (Dynamic World, The Global Forest Watch) since the year of observation is calculated for all observations since the baseline of the analysis (2016). More detailed field descriptions are provided in the respective REAMDE sheet.
• FilteringGuide shows the approach to filter each species observations for possible location errors based on broad landscape classes (Corine Land Cover classes are described in the according README sheet), as well as annotations for the necessity of land-use land-cover change analysis.

Geographic projection:

The observations are provided in LKS-92 / Latvia TM (default for all layers in HiQBioDiv).