HTTPS SSH
title

Species

subtitle

A web platform to find unkown relations hidden in millions of species occurrence records


Team member(s) names, roles and affiliations

Application Development Team

  • Raúl Sierra Alcocer, Eco-informatics Coordinator, CONABIO.
  • Juan M Barrios, Systems Architect, CONABIO.
  • Juan Carlos Salazar-Carrillo, Web Development Consultant, CONABIO.
  • Pedro Romero Martinez, Back-end Developer, CONABIO.

Scientific advisors

  • Christopher R. Stephens, Data Science Coordinator, C3, UNAM.
  • Constantino González-Salazar, Professor-Researcher, UAM.

R package developer

  • Enrique del Callejo Canal, PhD Student, IIMAS, UNAM.

Abstract and rationale

SPECIES is a spatial data mining web platform for the exploratation of biodiversity data. It has two models for analysis: ecological niche and ecological networks, both constructed from spatial correlations found in the databases. SPECIES integrates: data, a statistical engine, and an interactive visualization front end. Combined, these components provide an analysis tool that may guide ecologists toward new insights. SPECIES is optimized to support fast hypotheses prototyping and testing, analysing thousands of biotic and abiotic variables and presenting descriptive results to the user at different levels of detail. The rationale of this project is that, as of today, we can only explore GBIF occurrence records with questions that are explicit on the data fields, like, which are the records for Panthera onca, or which are records from Mexico, or from a specific project. Questions like \'which mammals are usually found near the jaguar?\', however, are not possible. We believe that the possibility of answering that type of questions opens up a new way of generating knowledge from GBIF and other, similar, biodiversity data repositories. We believe that given the size of GBIF data, which recently reached the 1 billion records, we need more tools that can extract patterns to help our communities to better exploit this world of information.

Operating instructions

Ecological Niche

Welcome to the Ecological Niche analysis section of SPECIES. An analysis worflow has the following steps.

  1. Target species configuration
    1. Choose a species: type the scientific name of the target species. As you write the system will give options. Click on the name you're looking for when it appears.
    2. Region: Choose the region that you'll use for your analysis, currently we have three countries (Colombia, Mexico and the USA) and one combined region (Mexico and the USA).
    3. Grid resolution: Choose the grid resolution that you'll use for your analysis, the grid defines the spatial partition to define co-occurrences of variables.
    4. Filter by date: Use the dates slider to define the time range of the records you want to include in your analysis.
    5. Include data that has no date: You can include data that has no associated observation date. The default is that they are included.
    6. Include fossil data: You can exclude fossil records of the chosen species. The default is that they are included.
    7. Histogram of observations: This histogram shows the occurrence records frequency by year. It updates when you modify the time interval with the slider.
    8. Apply filters: Once you have defined the filters for your analysis press this button to update the selection and the map.
    9. Occurrence map: In the first map you see records that will be considered for your analysis. Click on an occurrence to display its associated data.
  2. Analysis configuration
    1. Variable group: The next step is to define the covariates (niche predictors) that you will use in the model. In SPECIES you add them as groups of variables, you can add as many groups as you like.
    2. To add groups of species as covariates, choose the taxonomic level of the group of species you want to define and type the name of the taxon. For example, if you want to add all mammals, select the taxonomic level \'Class\', and then type \'Mammalia\'. Once the taxon is chosen the system shows the taxonomic subtree associated with the chosen taxon where you can choose to check/uncheck subgroups in order to include them or not in the analysis. Click on the plus sign button to create the covariates group.
    3. For raster data you will see a list of data sources, each data source contains a set of variables from that source. Choose the data source(s), and add them using the plus sign button, as in the taxa tab. You create multiple raster groups in this way
    4. Create variable group: Once you have chosen a variable group you should add them by clicking on the plus sign button.
    5. Parameters for niche analysis: You can adjust some parameters for the niche analysis.
    6. Minimum number of occurrences: Fixes the minimum number of cells that a covariate occupies to be included in the analysis.
    7. Run the analysis: Execute niche analysis with the chosen parameters.

Ecological Community

Welcome to the Ecological Community analysis section of SPECIES. Follow the steps to build a correlations network.

  1. Source variable group: The first step is to define source nodes that you will use in the model. Add the node groups one by one.
  2. Target variable group: The next step is to define target nodes.
  3. Variable types: In both groups, you can choose biotic and/or abiotic variables. Click on the corresponding tab of the variables you want to include.
    1. Taxonomic order: For biotic variables choose the taxonomic level of the group of species you want to define and write the name of the taxon. For example, Class: Mammals.
    2. Taxonomic tree: Once the taxon is chosen the system shows the taxonomic subtree associated with the chosen taxon where you can choose species from the subgroup.
    3. Raster Variables: Choose the following raster variables, they were processed to be used as covariates in your analysis.
  4. Create variable group: Once you have chosen a variable group you should add them by clicking on plus icon button.

  5. Minimum number of occurrences: Fixes the minimum number of cells that a variable should cover to be included in the analysis.
  6. Grid resolution: Choose the grid resolution that you'll use for your analysis, the grid defines the spatial partition to define co-occurrences of variables.
  7. Region: Choose the region that you'll use for your analysis.
  8. Run the analysis: Execute niche analysis with the chosen parameters.

Links to visuals

Link(s) to submission materials

License

All project components, SPECIES frontend, SNIB middleware, and speciesDBBuild, are licensed using the GNU Affero General Public License v3.0.