Taxonomic Coverage page mockup

[Matt Jones]

Thanks for the quick feedback.

Matthew Brooke wrote:
> Matt:
> 
> A few questions/discussion points regarding your Taxonomic Coverage page 
> mockup comments:
> 
> 1) "Not all taxa use the ranks listed in your table..."
> 
> a) Perumal and I talked about the approach you describe before coming up 
> with the current one, but the problem with a 2-col table is that every 
> taxonomic definition would need a separate table (and probably a 
> separate dialog), whereas perumal's mockup allows all the data to be 
> presented clearly and concisely in one place. Even if we did a 2 or 
> 3-col table, wouldn't we still need some kind of "summary" table like 
> this one (compare with party dialog & party table) ? We figured that the 
> clearest way to show the info was in this form, rather than a tree etc, 
> and it also allows all entries to be editable in place (except common 
> name - see below).

I understood the reason for the table being structured as proposed.  I 
was suggesting that upper level taxa (e.g., Kingdom, Phylum, Division) 
need not be displayed in the summary table at all.  Rather only the 
lowest most rank would be displayed in the summary.  Thus, you would 
still only need one table for all of the taxa. Here's an example:

For these taxa:
   Kingdom Animalia, ..., Class Crustacea, ..., Genus Bosmina, Species 
Bosmina longistrosis
   Kingdom Animalia, ..., Class Mammalia, ..., Genus Bos
   Kingdom Plantae, Division Magnoliophyta, ..., Genus Psychotria, 
Species Psychotria nervosa


the summary table would contain:

Rank                  Name
------                ----------------
Species               Bosmina longistrosis
Genus                 Bos
Species               Psychotria nervosa

Note the summary table does not display the upper level parent taxa. 
But that information would be shown when you edit the row.  For many, 
probably most, of the entries, people will not provide the parent taxa 
anyway.  The most common thing will be for them to only provide a list 
of the Species, but we do want to accomodate the addition of parent taxa 
as EML does.

> Would either of the following work?
> 
> i)  make the ranks (column headings) editable, but still use this 
> layout? If a single dataset contains different taxon ranking methods for 
> different entries (for example, if some rankings include Division, but 
> others include Phylum), then we may still need to go to multiple tables, 
> but would this be a much less-common scenario?
> 
> ii)  represent the info in tabular form without using col headings - so 
> for each entry we could indicate the rank alongside the name; eg:
> 
>  ... | Family: Flagellariaceae | Genus: Flagellaria | Species: 
> guineensis | ...etc
> 
> ("|" pipes used to show where column dividers fall)

These might work, but I still prefer my approach above.  I don't think 
the summary needs to display all of that information, especially as it 
often will not be provided anyway.

> 
> b) An additional comment:
> 
> "Common Names" can apply to any level of the taxonomy (I think?), but we 
> obviously have a problem representing that in the table as shown (unless 
> each entry has 2 lines?), so we resorted to showing an abbreviated 
> "surrogate" list in the last column (like we do for party etc), which 
> could be clicked/edited in some way to pop up a dialog showing which 
> common names correspond to which ranks. However, a drop-down list (as 
> shown in the mockup) probably isn't appropriate for the "common names" 
> column - since this implies a choice to be made by the user.
> 

Again, the common name probably need not be displayed at all in the 
summary.  And as you note, the common name can apply to any level in the 
hierarchy, so it could be confusing.  At most I would display the common 
name associated with the most precise taxon (so in the table above, you 
might add a 3rd column containing "wild coffee" for the Psychotria 
nervosa row (but I wouldn't add "Plant" to the common name, even though 
Psychotria is in Plantae.

> 
> 2) "When people use coded values in their metadata..."
> 
> The middle part of perumal's mockup shows a section entitled "Import 
> Taxon Data" which, if I understand correctly, is where a user would be 
> able to provide the taxon information from existing (or 
> yet-to-be-imported) data tables, as you describe

Yep, that works well for tables containing the list of names.  However, 
some tables will only contain species codes (e.g., PIRU), and if they 
have already defined that table, then the code definitions in EML will 
have the list of species names (e.g., PIRU -> Picea rubens), which I 
think should be easily transferred to the coverage element.  So I still 
think I'm describing a different case than is covered by Perumal's 
existing widget.  I'd like to extract the species list from the 
enumerated values definition in EML if the person took the time to type 
it into there.  This might most easily be done in the enumerated list 
widget where there might be a checkbox saying "These codes represent 
taxonomic species, so include the definitions as Species names in the 
Taxonomic Coverage section"  (wording needs to be less ackward, but you 
get the idea, right?)

Matt


> M
> 
> 
> Matt Jones wrote:
> 
>> Hi Perumal,
>>
>> That is a nice mockup.  A couple of issues come to mind.
>>
>> 1) Not all taxa use the ranks listed in your table.  For example, in 
>> plants they use 'Division' in place of 'Phylum'.  There are many of 
>> these variants, which is why in eml the system allows the user to 
>> specifiy both the rank and the name (rather than fixing the rank to a 
>> predefined set). So I would prefer a two column table containing 
>> "Rank" and "Name" for the most precise rank provided by the user, 
>> which might have values like "Species" and "Bosmina longistrosis".  
>> You could also list common names in this table if desired as a third 
>> column.
>>
>> 2) When people use coded values in their metadata (ie enumerations) 
>> they'll be defining the codes in terms of taxon names, and that same 
>> process can be used to fill out the taxon coverage info.  So that is 
>> another way through which they might provide the taxon information.
>>
>> Matt
>>
>>
>> Perumal Sambasivam wrote:
>>
>>> Hi All,
>>>
>>> For the taxonomic coverage page, I have a preliminary mockup (pls see 
>>> the attachment).
>>>
>>> It consists of -
>>> 1. A list for entering taxon names. The list consists of 7 columns for
>>> each level of classification. A user could add a species name and 
>>> specify
>>> all the names in the hierarchy. There is also a column for the common
>>> names.  It will be a dropdown box of all the common names of the whole
>>> hierarchy. If user wants to add a common name, a dialog box is presented
>>> with a customlist for each level and the user can add a common name 
>>> to any
>>> level.
>>>
>>> 2. The user can import the taxon data from a particular table. The 
>>> column values will be read, an ITIS lookup will be done for each 
>>> value to determine its level and values in the hierarchy above it and 
>>> the values will be added to the above list. The user can edit any 
>>> value, if needed.
>>>
>>> 3. The user can also optionally specify a new Classification system 
>>> and identifier.
>>>
>>>
>>> Please let me know your comments/feedback on this.
>>>
>>> Thanks!
>>> -Perumal
>>>
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------
>>> Perumal Sambasivam
>>>
>>> National Center for Ecological Analysis and Synthesis (NCEAS)
>>> University of California Santa Barbara
>>> ------------------------------------------------------------------------
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>

-- 
-------------------------------------------------------------------
Matt Jones                                     jones at nceas.ucsb.edu
http://www.nceas.ucsb.edu/    Fax: 425-920-2439    Ph: 907-789-0496
National Center for Ecological Analysis and Synthesis (NCEAS)
University of California Santa Barbara
Interested in ecological informatics? http://www.ecoinformatics.org
-------------------------------------------------------------------