MyWoRMS
In CRITTERBASE each used taxon is tested against WoRMS. Since a query at WoRMS takes a long time, WoRMS is not queried
directly but a local extended buffered instance of it, the so-called MyWoRMS.
Buffer
All requests to MyWoRMS are buffered. If a request is already found, it is returned from the buffer without asking WoRMS
again. Two different requests are distinguished.
Queries by Aphia Id
The queries are guided by the Aphia Id and return the taxon found.
In the buffer these taxa are stored together with the search key (here the Aphia Id).
Each taxon also stores the date of its query or creation in MyWoRMS.
The following image shows the result of the manual search in MyWoRMS for the Aphia Id 141433.
Queries by name
Here the requests are stored by name.
This feature is especially useful for TaxaPolishing, because there you can search for
taxa by name.
However, such a name query can be performed in different ways with multiple answer options.
Please see the corresponding chapter about TaxaPolishing for more details.
The following is also important to mention when searching for a name.
There are three types of name search.
Search types
- normal
- fuzzy (generic)
- fuzzy (special)
With the normal search, exactly the search word is searched for and there is always only one answer.
This search is not suitable because no taxon will be found if there are spelling mistakes in the search query.
The generic fuzzy search (SQL: LIKE)
fuzzily searches in WoRMS data.
Thus, typos are detected in the search query and many answer options are returned.
The special fuzzy search query (TAXAMATCH by
Tony Rees)
also fuzzily searches the WoRMS dataset and returns multiple possible answers if necessary.
Since WoRMS also returns unaccepted taxa as possible answers for both searches, the algorithm in CRITTERBASE was
extended externally so that each unaccepted search result is followed up to an accepted one.
If no accepted taxon can be found, the answer is discarded.
If no answer option remains, the original answer is returned to the user by WoRMS.
By the way, all taxa queried by WoRMS in this way are also buffered in MyWoRMS.
The following image shows the result of a manual fuzzy search in MyWoRMS for the name Abra
.
The following image shows an overview of all bufferd name queries in a MyWoRMS buffer.
Buffer Quality
The most important criterion for the quality of a buffer is the age of the stored requests.
The following image shows the main status information on MyWoRMS.
Maximum taxon age in MyWoRMS is 30 days
Each taxon in MyWoRMS must not be older than 30 days.
Then it is automatically retrieved a new one from WoRMS when it is used again (request to MyWoRMS).
Thus, the quality of all controls in CRITTERBASE can be ensured by a current locale MyWoRMS dataset.
De-Ageing of taxa in MyWoRMS
Since each taxon remembers the time of its creation in MyWoRMS, it is possible to get an overview of the age of the
whole MyWoRMS system at any time.
If the buffer becomes too old, the system warns you and you can de-age all taxa in advance.
This is done fluently staggered by taxon age group.
The oldest age groups are de-aged first.
The de-aging process can be interrupted and continued by the user at any time.
This manual de-aging ensures that not all obsolete taxa have to be updated during an upcoming long-term import process.
This undertaking can therefore be carried out manually in advance at your leisure, if required.
The following image shows an overview of all misfit taxa in a MyWoRMS buffer.
Misfts in MyWoRMS
Furthermore, it shows how many fismits are in the MyWORM system.
Taxa are counted as misfits, which are described in the following chapter Misleading Accepted Aphia Ids and
Taxon problems.
The following image shows the result of De-Aging process of the MyWoRMS buffer.
Backup and restore
Furthermore, the entire MyWoRMS instance can be saved per snapshot and restored on other computers.
Thus, costly and time-consuming rebuilds of MyWoRMS on multiple computers within the 30-day update window are easily possible.
Direct query
Direct queries with all the advantages of the taxon tests of the CRITTERBASE system can also be performed directly.
This also has the advantage that each taxon queried in this way is immediately in the MyWoRMS buffer.
It is also a good way to investigate misfits more precisely in the MyWoRMS buffer.
Especially the deficits of the WoRMS search, e.g. not recognizing the errors described below and the automatic
determination of a really accepted taxon in case of an unaccepted one is only possible in this way.
WoRMS cannot do this out of the box per web page.
Representation of taxa in MyWoRMS
Textual display of taxa
Taxa are displayed in two different modes. Normal mode is active when the taxon contains no errors. Debug
mode is selected when taxon problems are detected.
Normal mode
The presentation follows the following form:
Accepted Taxon and Taxon in Limbo
Format:
AphiaId Scientificname (Authority) [Rank]
Examples:
130503 Ophelina cylindricaudata (Hansen, 1879) [Species]
1380335 Lyonsia densecostata (W. H. Turton, 1932) [Species] <limbo: taxon inquirendum>
Unaccepted Taxon
Format:
AphiaId {Accepted AphiaId} Scientificname {Accepted Scientificname} (Authority) [Rank] <Status>i
Example:
131484 {868065} Axionice maculata {Pista maculata} (Dalyell, 1853) {Dalyell, 1853} <unaccepted: superseded
subsequent combination>>
Debug mode
The presentation follows the following form:
Format:
AphiaId (Valid AphiaId) {Accepted AphiaId} Names: ScientificName (Valid ScientificName) {Accepted
Scientificname} Authority: Authority (Valid Authority) (Accepted Authority) [Rank] <Status>
Example:
850475 (861599) {861599} Names: (Thyasira beui) {Thyasira beui} Authority: (Amano, Little, K.A.
Campbell, R.G. Jenkins Saether, 2015) {Amano, Little, K.A. Campbell, R.G. Jenkins Saether, 2015} <unaccepted>
[status:deleted, unacceptreason:duplicate (Bouchet, Philippe)>]
Taxon States: accepted / unaccepted / limbo
A taxon can have one of the following states with respect to acceptance. The states are mutually exclusive.
- accepted
- unaccepted
- limbo
Unfortunately, the states are not quite clearly classifiable into the two states accepted and unaccepted
on the basis of WoRMS.
Because of this the third state called limbo is located between the other two. A taxon in state in Limbo
is on its way from accepted to unaccepted, but has not yet arrived there, or perhaps will never.
The following parameters determine in which state a taxon is.
Except for the parameter Accepted AphiaId, which CRITTERBASE determines automatically (Aphia Id of the last valid
taxon in a taxon validity chain), all Parameters come directly from WoRMS.
For the sake of completeness, it should be noted that the Unaccept Reason parameter has no influence on the state.
Only a configuration warning is issued if an Unaccept Reason was specified for an accepted taxon.
- AphiaID (WoRMS)
- Valid AphiaId (WoRMS)
- Accepted AphiaId (CB)
- Status (WorMS)
- Unaccept Reason (WorMS)
The determination of a state of a taxon is done using the following logical state matrix.
The *-sign means that all inputs are disregarded.
State |
AphiaID=Valid AphiaId=Accepted AphiaId |
|
Status |
|
Unaccept Reason |
accepted |
true |
and |
="accepted" |
and |
* |
unaccepted |
false |
or |
="unaccepted" |
and |
* |
limbo |
true |
and |
!="accepted" |
and |
* |
Misfits
Taxa with Misleading Accepted AphiaIds
In addition to the configuraiton problems of a taxon discussed below, there exists a condition that is also undesirable.
It basically represents a retroactive misconfiguration of a taxon that is itself unaccepted and points to an accepted
taxon with its Accepted AphiaId.
At some point, this very taxon changes its status from accepted to unaccepted or in limbo.
The feedback about the resulting misconfiguration of the first does not happen in WoRMS.
CRITTERBASE analyzes and detects such misconfigurations.
Taxa with other Configuration Problems
There are three configuration problems with taxa itself that are detected by CRITTERBASE and also lead to the debug
display of the taxon.
Problems with Aphia Ids
A problem is recognized if one of the following rules applies to a taxon:
- No AphiaId given
- AphiaId given, AcceptedAphiaId given and no ValidAphiaId given
- AphiaId = AcceptedAphiaId != ValidAphiaId
Problems with Names
A problem is recognized if one of the following rules applies to a taxon:
- No ScientificName given
- Taxon is accepted and no AcceptedScientificName given
- Taxon is accepted or in Limbo and it does not apply: ScientificName = ValidScientificName
= AcceptedScientificName
Problems with Status
A problem is recognized if one of the following rules applies to a taxon:
- Only one status per taxon is allowed: accepted, unaccepted, in limbo
- Taxon is accepted and no AcceptedScientificName given
- Taxon is accepted and an unaccept reason is given