Milestones

Database milestones

BioDeepTime is drawing information from both static and dynamically-updated databases. Due to the heterogeneity of the constituent data, BioDeepTime is compiled manually in semi-regular intervals. These data products are versioned semantically, and are deposited separately - including the code that was used to build the version. These static data products were chosen to enhance reproducibility and tracability of results based on the database.

Versioning

The planned versioning framework is:

  • First-level: Large structural changes and major addition of sources.
  • Second-level: Corrections and additions of new data. Smaller structural changes.

Contents

List of known issues

MARBEN

  • Missing species and genus columns

General

  • References need manual review especially bibtex. Character encoding is still a recurring problem. Some references entities represent multiple references.

Change log

BioDeepTime v1.0 [2023-07-12]

DOI

Added

  • new field samples.totalCount that represents the sample size in the case of count data, rather than the target sampling effort
  • Three new abundanceUnit categories: "biomass cover", "biomass weight", "biomass volume"
  • bibtex handles to the bibtex column of the refs table. The candidate bibtex entries are in the refs.bib file.

Changed

  • The word occurrence was systmatically replaced with record. The occurrences table was renamed to records, its primary key from occID to recordID.
  • The timeUnits table was renamed to timeOriginalUnits for better consistency. Consequently, the the fields timeUnitID and timeUnit were rename to timeOriginalUnitID and timeOriginalUnit, respectively.
  • The ranks table was renamed to analyzedRanks for better consistency.
  • Reference entries are forced into UTF-8 encoding
BioTime
  • Omitted studies 39 and 217 due to potentially erroneous entries
  • Added biomass data to where there were no abundanceUnits earlier
  • Time series taxonomic/environment groups are added
  • biomass values replaced abundance values in cases when count data was given as 0, but biomass was valid
Neotoma
  • sample sum count is moved to samples.totalCount from samples.samplingEffort. Accordingly samplingEfforType is consistently set to NA.
  • Neotoma references were split, multiple refs per samples are now properly indicated
  • Changed the taxon group to Plants
Triton
  • New version is used now - indicated to be released soon as Triton 2
  • Fixed issues where all abundance values were relative abundances, even when count was indicated.
  • The sample sum is now recorded in samples.totalCount and not in samplingEffort. In case where samples reflect normalization for 1 gramm, the value 1 is now recorded in samplingEffort with a samplingEffortType of “g”.
SedTraps
  • added missing samplingEffortType (all are m^2)
MARBEN
  • added samplingEffort and samplingEffortType values from processing
  • total count of count-type data
  • added missing reason: "Community analysis"
  • added reference to MARBEN
PBDB
  • added total count of count-type data
Direct uploads
  • coccolithophore data: moved total count from samplingEffort to totalCount

Removed

BioTime
  • Studies 39 and 217 were removed due to quality reasons.

BioDeepTime v0.6 [2023-01-17]

DOI

Added

  • The Geobiodiversity Database - the Fenxiang section

Changed

  • moved the coccolith data to “Direct uploads”

Deleted

  • occurrenceTypeID column and occurenceType table

Acknowledged

  • The two Neptune time series that have mixed taxonomy represent true data.