VegBank

A few merges of references or deletions of duplicates still needs to be made,
otherwise, this is completely successfully.

The following SQL needs to be run against vegbank to accomplish the above.
Then this bug goes back to Lee so that he commits all this into cvs so that it
will be done next time.

--------------------
DELETE FROM referenceContributor where reference_ID = 26;
DELETE FROM reference where reference_ID = 2;
DELETE FROM reference where reference_ID = 4;
DELETE FROM reference where reference_ID = 26;

UPDATE reference SET shortName = 'EcoArt 2002'   WHERE (((reference_ID)=32));
UPDATE reference SET shortName = 'USDA Plants 2002' WHERE ((reference_ID)=33);

sql committed against vegbank 05Jun2003 4:58 PDT. Successfully changed
references to the way they should be. This done by lee.

THIS BUG HAS WEIRD ORDER OF COMMENTS DUE TO ECOINFO'S CLOCK BEING OFF on
05JUN2003.

email from Bob was actually on this date:
Sun, 27 Apr 2003 07:13:28 -0400 (EDT) not the date mentioned in the earlier comment

Journals have been captured! The only remaining questions follow here:
-------QUESTIONS---(These journals not yet included in VegBank)-------------
ANN BOT-LONDON =?? ANNALS OF BOTANY is in Oxford, not London
CONSERV BIOLDODO =could not find bioldodo anywhere, not even google

-------THINGS I THINK I FOUND, BUT INCLUDE HERE TO DOUBLE CHECK--------
(these included in Preload for VegBank)
AUST J ECOL =found on web, AUSTRAL ECOLOGY?? - new name
RES POPUL ECOL =Population Ecology?? formerly Researches on Population Ecology
GREAT BASIN NAT =found on web, BYU Journal?
CASTANEA =found on web, no ISSN - emailed them to ask.
ANNU REV PLANT PHYS=web: Annual Review of Plant Biology, Formerly The Annual
Review of Plant Physiology & Plant Molecular Biology

Bob, if any of these are wrong (below) or you know more about the ones above,
let me know and I can fix/add.

Journal names were from ISI

ANN BOT-LONDON =?? ANNALS OF BOTANY is in Oxford, not London
Annals of Botany -- recenly Oxford, but earlier London, me thinks
CONSERV BIOLDODO
Conservation Biology
AUST J ECOL =found on web, AUSTRAL ECOLOGY?? - new name
No, the older Australian Journal of Ecology
RES POPUL ECOL =Population Ecology??
formerly Researches Population Ecol.
GREAT BASIN NAT =found on web, BYU Journal?
Great Bsin Naturalist
CASTANEA =found on web, no ISSN - emailed them to ask.
OK - in that I have published in this backward little journal
ANNU REV PLANT PHYS=web Annual Review of Plant Biology
The former The Annual Review of Plant Physiology & Plant Molecular Biology

Journals are done. Annals of Botany added with abbrev: "ANN BOT-LONDON" and
another reference to Conservation Biology was made using the (weird)
abbreviation "CONSERV BIOLDODO" - is this correct or do we only want one record
for conservation biology, with the abbreviation "CONSERV BIOL" which was also in
the list and an unambiguous match?

"AUST J ECOL" points now to Australian Journal of Ecology, with the same issn as
Austral Ecology, the same journal, but newer name.

Still need references, methods.

From an ancient .doc from Bob to John Harris regards project 41 of the NCVS
loading, I glean the following regarding references:
(I can add these, but I think #3 needs some editing to fit our new data.
Furthermore, see the note at the end which dislikes our (once again) current
approach about splitting up authors. Maybe we need to retain a reference field
which stores the names in exact order, spelling, abbrev. etc.?
------------
3. citation

For now I have only three references that need to be cited, as follows

Entry #1
citAuthors: Robert K. Peet, Thomas R. Wentworth & Peter S. White
citTitle: A flexible, multipurpose method of recording vegetation composition
and structure
citPubDate: 1998
citSeriesName: Castanea
citIssueIdentification: 63(3)
citPage: 262-274

Entry #2
citAuthors: J. Braun-Blanquet
citTitle: Pflanzensoziologie: Grundzüge der Vegetationskunde.
citPubDate: 1928
citSeriesName: Biologische Studienbücher 7
citPublisher: Springer
citLocation: Berlin

Entry #3
citAuthors: The Association for Biodiversity Information
citTitle: Ecology Access Reporting Tool. International classification of
ecological communities. Natural Heritage Central Databases.
citEdition: 2.0
citEditionDate: September 13, 2000
citPage:
citPublisher: The Association for Biodiversity Information
citLocation: Arlington, Virginia

Notes:
3.1  At present we pull author name out of citation and placed that in
citationContributor.  I think this a mistake.  Librarians and other
bibliographically compulsive people want exact the spelling of author name(s) as
in the text.  We do not allow for this when we reference back to vegPlotParty
(unless we want to track all spellings and abbreviations).
3.2  Note that I have adjusted some of the other fields of citation to fit
present needs.

From the same .doc as above, cover method and cover index are as follows. Do we
need others?
-------------------------------
5. coverMethod

Entry #1: coverType = Braun-Blanquet; citation_FK=Braun-Blanquet 1928 (2)
Entry #2: coverType = Carolina Vegetation Survey; citation_FK=Peet et al 1998 (1)

7. coverIndex

Below I provide entries for coverIndex for two scales. The fields are
CoverMethod_ID (BB=1, CVS=2?), indexCode, lowerLimit, upperLimit, indexDescription

BB r 0.01 0.05 One or a few individuals
BB + 0.05 0.1 Occasional and < 5%
BB 1 0.1 5 Abundant with very low cover, or less abundant
with higher cover, always <5%
BB 2 5 25 very abundant & <5%, or 5-25%
BB 3 25 50 25-50%
BB 4 50 75 50-75%
BB 5 75 100 75-100%
CVS 1 + 0.1 Trace
CVS 2 0.1 1 0-1%
CVS 3 1 2 1-2%
CVS 4 2 5 2-5%
CVS 5 5 10 5-10%
CVS 6 10 25 10-25%
CVS 7 25 50 25-50%
CVS 8 50 75 50-75%
CVS 9 75 95 75-95%
CVS 10 95 100 95-100%

From same .doc as above, I get the following stratum Types for NCVS. Need
stratumTypes for NPS plots, too. Others? BB? Also need name of stratumMethod
for NCVS (Carolina Vegetation Survey) and a description (???).
---------
8. stratumType

Below I populate stratumType with 5 entries for the fields stratumName and
stratumDescription.

NCVS-1 Foliage <0.5m high
NCVS-2 Foliage 0.5-6m high
NCVS-3 Foliage 6-15m high
NCVS-4 Foliage 15-35m high
NCVS-5 Foliage >35m high

In the above comments, I've now at least scratched the surface on many
preloading tables. The only ones that we might further consider are:
---------
userDefined (seems silly to prepopulate this, but we may want to)
party (maybe add VegBank group - would be good to have data in there)
project (doubtful)
---------

Otherwise, we will add some plants and comms and plots to the system before
release, but this will be dealt with in a different manner.

1. We should add the major covervalue systems supported by TurboVeg (see
attachment). We should probably insert all of these to maximize compatability
wit TurboVeg.

2. Regards authors names for references. I had thought we had worked out way
through all of this. The reason we lumped multiple parts into
referenceParty:givenName is so that we can optimize the match with the format
in the reference. I would want to maintain all spellings nad formats used in
different articles. The reason for the referenceContributor:position field is
to keep track of the order of the contributors.

3. I am not sure what Mike's comment "Note that I have adjusted some of the
other fields of citation to fit present needs" refers to, but would very much
like to know the details.

We're making good progress with this bug!

CONSERV BIOLDODO has gone the way of the dodo (sorry, that's not in good taste).
DODO is still there, as is CONSERV BIOL.

I will do the best I can with naming TurboVeg stratumMethods. Should we use
TurboVeg as a reference, at least until better references can be found? I also
assume raunki‘r is meant to be Raunkièr?

Bob's comment from before:
---------------------------

2. Regards authors names for references. I had thought we had worked out way
through all of this. The reason we lumped multiple parts into
referenceParty:givenName is so that we can optimize the match with the format
in the reference. I would want to maintain all spellings nad formats used in
different articles. The reason for the referenceContributor:position field is
to keep track of the order of the contributors.

So long as the authors always are listed givenName(s) surname, givenname(s)
surname, then we're ok. If sometimes they are listed surname, givenname(s);
surname2, givenname(s)2, then there's not a way with our model to know that we
should recreate the latter. I don't know if this happens or not. I just
included the note as I discovered it and wondered if this was considered when we
changed the model again. It seems that it was considered and the givenName
field containing all given names was the solution.

Bob's comment from before:
---------------------------

3. I am not sure what Mike's comment "Note that I have adjusted some of the
other fields of citation to fit present needs" refers to, but would very much
like to know the details.

This comment is verbatim from the Word doc to John Harris, which I will attach
to this file, saved as an html, I guess. Whatever will work.

Also, I grabbed the stratumTypes from the PLOTS database,
stratumIndex stratumName
T1 Emergent
T2 Canopy
T3 Sub-Canopy
S1 Tall Shrub
S2 Short Shrub
S3 Dwarf Shrub
H Herbaceous
N Nonvascular
V Vine/Liana
E Epiphyte

1. Regards References
Raunkiaer, C. 1934. The life forms of plants and statistical plant
geography. Clarendon, Oxford. 632 pp.

Some are simply intuitive and we can reference TurboVeg (such as
presence/absence)

For the others, you might email either
Joop Schaminée (J.H.J.Schaminee@Alterra.wag-ur.nl) or
Stephan Hennekens (S.M.HENNEKENS@IBN.DLO.NL)

2. I think we should leave the reference model alone for now. The author info
is close enough, until we have an expert in the field advise us otherwise.

time has been restored?

Check in the method that was used to prepopulate the database.

Make sure Castanea's ISSN has been added:
Journal ID

Journal Name

Abbreviation

ISSN
6 CASTANEA CASTANEA 0008-7475

where to put the stuff in cvs:
I created a little shell script to allow ant to call the preloading of
the db from a target it is now in cvs.
vegbank/bin/loaddb.sh

I think vegbank/build.xml is a good place for your new targets. Bellow
is the bottom of the build file where I have placed some db related
targets. The "<exec executable=" shows how to call a shell script from
ant.

The directory ${build.sql.src.dir} is where auto-generated sql should
go.

#################################################################
<!-- =================================================================
-->
<!-- Database related targets
-->
<!-- =================================================================
-->

<target name="db_generate_sql"
description="==> Generate the sql build scripts"
depends="prepare">

<xslt in="${docs.dir}/xml/db_model_vegbank.xml" 
                  out="${build.sql.src.dir}/vegbank.sql" 
                  style="${xsl.src.dir}/VegBankModel2SQL.xsl">
            </xslt>
  </target>

<target name="db_rebuild_vegbank"
description="==> Rebuild the vegbank Database"
depends="db_generate_sql">
<!-- DROP and CREATE Database -->
<exec executable="${bin.dir}/rebuilddb.sh">
<arg value="${db.name}"/>
<arg value="${build.sql.src.dir}/vegbank.sql"/>
<arg value="veg_plot/sql/makePlotSummaryTables_postgres.sql"/>
</exec>
</target>

This is complete in the sense that the database is preloaded successfully. It
is not yet documented completely how we did this, nor how one would repeat the
preload. I worked on this a bit, but ran into lack of sufficient memory in
transforming large XML docs into SQL files which contain the insert commands to
preload vegbank. So, it is waiting on a solution to the memory problem.
(initially, the XML was transformed using XMLSpy on my machine, which had enough
memory allocated).

in src/sql for now.

no longer need to document in cvs: is part of the root db which will need
preserving with accessionCodes, etc.

changed from components that are to be deleted to "misc" so that bugs don't get
deleted with component. Sorry for all the email.

Original Bugzilla ID was 1077