Skip to content

Home

Jump to bottom
Stan Blum edited this page Nov 13, 2021 · 11 revisions

Welcome to the material-sample wiki! Please feel free to add to and edit pages here.

The main challenge facing primary biodiversity data publishing and integration is to agree on how publishers should encode and how aggregators and end users should interpret data about:

  1. Organism
  2. Occurrence, and
  3. Material-sample (all the subtypes)

Some simple definitions taken from wikipedia or dictionaries:

  • organism: wikipedia an organism (from Greek: organismos) is any organic, living system that functions as an individual entity.

  • occurrence (defining a biological occurrence): [Not from dictionary or Wikipedia] The existence of an organism at a place and time.
    The only restriction we might put on this is that occurrence could/should be categorized as natural (occurrence in nature) versus human facilitated. This is controversial (both sides of this have been expressed in the discussion), needs resolution, and our definition should include clarification of this.

  • material-sample: wikipedia: a sample is a limited quantity of something [or in BDI usage a group of organisms, a single organism, or a part of an organism, or parts of multiple organisms, including free macro-molecules] that is intended to be similar to and represent a larger amount of that thing(s); i.e., local population(s) of organisms. The things could be countable objects such as individual items available as units for sale, or an uncountable material.

[Can we defer talking about observations (bco:information-artifacts) in this analysis? Is it true that wherever an observation could be considered our scenarios, a material sample could be also collected in a comparable scenario? We can check later whether the absence of a material-sample should change how data are published and interpreted in ways that are not obvious.]

Class - Properties

Where do (should) properties attach to these classes? [For convenience, I've used informal language to denote property concepts rather than DwC terms. We can substitute DwC terms later.]

  • Organism [Org]: Taxonomic Identification (invariant), caste (social insects), sex (can change over time in some taxa), age-class (changes over time, so better a property of Occurrence)

  • Occurrence [Occ]= { [Organism] + [Event] } : [Date], [Time], [Collector(s)], [Method] + [Locality] LocalityDescription, Latitude, Longitude, Country, State-Prov, County, etc.

  • Material-sample [MatSam]: parts of organism, preparation (method and materials), storage, disposition, history (loan, exhibit)

Encoding classes represented or inferred

How does the publisher communicate to the aggregator and downstream users what exactly is represented in a record, and if multiple records are published for a given occurrence or organism, which record is the best representation of the thing the aggregator or end user is interested in?

Scenarios

  1. The trivial case occurs with a discipline like herpetology, where an organism is collected and preserved whole, and represented that way in the collection catalog. (Comparable cases occur in virtually every other collection discipline.)

    1 organism : 1 occurrence : 1 material-sample

    • [Org] dwc:scientificName, [Occ] dwc:occurrenceID, dwc:eventDate, dwc:recordedBy, dwc:locality, [MatSam] materialSampleType (Preserved Specimen), dwc:preparations: "whole animal (EtOH)"

  2. a salamander (organism) is collected, a tissue sample taken and placed in 95% EtOH and frozen in liquid nitrogen (LN2). There are two material samples, the "whole" specimen and the tissue sample. (Typically the whole specimen is also considered the organism and occurrence; the tissue-sample typically gets another identifier, at least a subnumber, so that it's trackable independently of the whole specimen.)

    1 organism : 1 occurrence : 2 material-samples

    Main-Record

    • [Org] dwc:scientificName, [Occ] dwc:occurrenceID, dwc:eventDate, dwc:recordedBy, dwc:locality, [MatSam] materialSampleType (Preserved Specimen), dwc:preparations: "whole animal (EtOH) | tissue (EtOH)"

      Related-Many-Records

      • dwc:occurrenceID, dwc:materialSampleID, dwc:preparations: "whole animal (EtOH)"
      • dwc:occurrenceID, dwc:materialSampleID, dwc:preparations: "tissue (EtOH)", ggbn?:storageRegime: "ultracold"

  3. a mouse is collected, a tissue sample taken and the whole specimen is prepared into a standard skin and skull. This is essentially the same as #1 except that the preparations are obviously not the whole specimen.

    1 organism : 1 occurrence : 3 material-samples

    • [Org] dwc:scientificName, [Occ] dwc:occurrenceID, dwc:eventDate, dwc:recordedBy, dwc:locality, [MatSam] dwc:preparations: "Skull" | "Skin" | tissue (EtOH)"
      • dwc:occurrenceID, dwc:preparations: "Skin"
      • dwc:occurrenceID, dwc:preparations: "Skull"
      • dwc:occurrenceID, dwc:preparations: "tissue sample: 95% EtOH stored in -80 deg C"

  4. three cuttings are taken from a shrub, one kept, two distributed to other herbaria

    1 organism : 1 occurrence : 3 material-samples (all with different institution/collection codes and accession numbers)

     The disipline-specific practice is to identify the organism:occurrence by the "CollectorsNumber".

 [Still working on the encoding scheme for this scenario.]
 {
 * instCode, collCode, catalogNumber, collectorsNumber, dwc:occurrenceID 
 * instCode, collCode, catalogNumber, collectorsNumber, dwc:occurrenceID
 * instCode, collCode, catalogNumber, collectorsNumber, dwc:occurrenceID
 }

  5. same plant collected repeatedly at different times of year to get buds, leaves, flowers, fruits

     - 1 organism : 4 occurrences : 4 material-samples
     - 4 separate herbarium sheets, each with a different accession number.

         Note, the occurrences are different times, but the same locality

  6. Endangered Hawaiian bird captured, banded, blood drawn to screen for malaria (viruses, etc.), then released, recaptured, and blood drawn again.

    1 organism : 2 occurrences : 2 material-samples [several subsamples could be derived from these, each with subsampling who, when, what, how data]

  7. [more examples where an organism is sampled or observed repeatedly...]

Clone this wiki locally