Migrate Biological Associations data into TaxonWorks

[weevil-see]

Can TaxonWorks conveniently store biological associations data which looks roughly like this:
1 parasite name
2 host name
3 various fields with information which define the association (which organ is used by the larva, larval biology (gall, vagrant, boring etc), perhaps also information on the occurence (in a darwin core sense) from which the information was derived
4 source citation

[mjy]

Yes, we have very rich model for storing that kind of data (and it will become even richer this year with reference to instance of AnatomicalParts), and a variety of nice interfaces for adding data. There is not currently a batch upload though.

[weevil-see]

Yes, we have very rich model for storing that kind of data (and it will become even richer this year with reference to instance of AnatomicalParts), and a variety of nice interfaces for adding data. There is not currently a batch upload though.

So you'd recommend to add data directly in TaxonWorks? I guess export of data is easy? Can this kind of data easily be accessed from TaxonPages?

[mjy]

So you'd recommend to add data directly in TaxonWorks?

Yes? Depends though- if you have a GLOBI level of data we'd want to work on an importer. We have a full DSL that you could script around, and have us run that script on production.

I guess export of data is easy?

2(?) years ago I developed multiple exports, including a DwC extension that is live. Most replicate/model after the kind of data GLOBI ingests.

Can this kind of data easily be accessed from TaxonPages?

Not yet, however this is coming "Any Day Now(tm)". We're essentially waiting to build an cached index of the data to server them very quickly (they are stored in very relational format).

[weevil-see]

As far as I can see, GLOBI is recording much less data than we would like to have:
For us it is important not only to pair both interactors, but to define the relationship. What plant organs are eaten by the larva? Some develop in flower buds, some in the stem, some are leafminers, etc. Adults often have broader host ranges than their larvae.
We also want to categorize information from published sources by how much evidence they provide. If a taxonomic catalogue is just listing a plant as host plant for a weevil without further information, this is less credible than a paper which describes how the species was reared from larva in a flower head of plant xy.

[mjy]

What plant organs are eaten by the larva?

This is the AnatomicalPart extension coming this year. The ability to bind BiologicalAssociations to AssertedDistributions is also coming within the year. "This larva localized_to this Taxon <your_relationship> ... CollectionObject" in Germany.

We also want to categorize information from published sources

We do this very well. You can cite any triplet and multi-cite any graph of triplets.

[weevil-see]

The "AssertedDistribution" you mentioned is also something I have thought about earlier, but I wasn't aware that there is the DwC term. It's clear that host ranges are not the same across the geographical range of species.

So there are two approaches on how to collect and maintain the data:

1. A database filled with raw data. It would contain occurences with biological associations only.
2. A database to make "aggregated knowledge" accessible and searchable. This type of database would provide information is a format which is easily understandable by users and can be used to direct fieldwork. This database should indeed contain "AssertedDistribution" of some sort.

My idea was to collect raw data in the sense of 1), which could then be aggregated into a database like 2).

One collaborator suggested earlier that it would not be enough to give information on Distribution on the scale of Country, as there are weevils even within Germany (comparably small country) which apparently have different biological associations in different parts of Germany.