The LEDA Traitbase Web Query Manual

Document Sample
The LEDA Traitbase Web Query Manual Powered By Docstoc
					The LEDA Traitbase Web Query Manual
Version: 12th of July, 2006

The LEDA web query tool has been designed for experienced users to access the data stored within
the LEDA Traitbase. It offers the possibility of querying data at the raw data level, a pregaggregated
data level and an aggregated data level. The LEDA species fact sheet application offers a simpler
possibility to access data. It will output a single value per species and trait.


Purpose of the Web query tool
The LEDA web query tool has been designed for retrieving data from the LEDA Traitbase at
different aggregation levels. Query results are displayed on standard web browsers and can be sent
to the user via e-mail.

Data Organisation
In order to perform queries, you need a basic knowledge about the way data is structured within the
LEDA Traitbase. Basically, we keep data for each trait in a separate table. Each record worth of
trait data may come from one of the LEDA partners or it may be calculated from a number of these
records. The records delivered by the LEDA partners may either contain information gathered by
the partner or taken from other databases or taken from literature. Each trait record may contain
information about a single measurement or observation or it may contain information about an
entire experiment or from multiple observations. In the former case, the record is at the raw data
level. In the latter case, the record is at the preaggregated data level.
Queries either retrieve portions of records described above or they may result in output of
aggregated data calculated from those records.

All you need to use the web query tool besides a computer connected to the Internet is a
contemporary Web Browser with both Java Script and Cookies enabled. Without cookies, the query
tool wouldn't be able to distinguish between different sessions and could not keep track of your
work's progress. Without Java Script some functions just do not work. E.g. pressing the Execute
query button would have no effect.
The query tool has been tested with the following browsers: Opera (from version 6), Mozilla
(version 1.7 and higher), Mozilla Firefox (version 1.0.7 and higher), Internet Explorer (version 6
and 7), Netscape Navigator (version 7.1).
Please never use your browser's back button in order to navigate within the web query tool. This
might confuse the tool and produce wrong query results.

Accessing the query tool
In order to access the query tool, you first have to register using the LEDA portal accessible via the
URL If you are already a registered user, you have to log in by
giving the system your user name and the password you selected during registration and
subsequently clicking the Login button.
After logging in, on the menu bar to the left, first click the Applications entry and then the Query
item which will appear.
The LEDA portal page on which these actions take place is depicted in figure 1. Depending upon
your access rights and the portal version you may see more entries appear under the Applications
menu entry.
After clicking the Query menu entry, the LEDA web query application's main screen depicted in
figure 2 will be shown.

                                       Figure 1: The LEDA portal.

                                Figure 2: The LEDA web query main screen.

Composing and executing queries
To start composing a new query, click on the Query Composition Tab. This will open the query
composition view (see figure 3). A simple query can be composed in three steps:
   (1) Select traits from the select traits dialog at the view's upper left section. Traits can be
       selected from the drop down list and then added by clicking the Add Trait button. It will
       then appear in the list of selected traits below the dropdown. In order to remove a trait from
       this list, click the remove trait button at the side of the trait name. Please note that raw data
       queries can only be performed for single traits.
   (2) Select attributes from the select attributes dialog at the view's upper right section. An
       attribute is a table column describing a trait characteristic (e.g. average leaf size) or the
       context under which the characteristic was observed (e.g. species name, country, number of
       replicates involved in the experiment). The drop down list contains all attributes available
       for the chosen traits. Attribute selection is done in the same way as trait selection. For an
       explanation of names appearing in the list, please see the Appendix to this document.
   (3) Order the list of attributes to suit your needs in the sample query section at the view's lower
       middle section. The main part of this section displays the headings of the query result table.
       You may move a column (an attribute) by selecting it (point in the circle field above the
       column name and click it) and then using the buttons at the top of the sample query section.

                                    Figure 3: Query composition View
After composing a query you may tailor it to your needs by selecting the columns which will be
used to order the query results (click into the ordered checkboxes) and by checking the null values
allowed box, if needed. If null values allowed is not checked, a query will only return rows that
have values for each cell. This might be helpful for data analysis purposes, but it will also
substantially reduce the number of results returned, in general.
After completing those settings, click the execute query button. After the database query is finished,
the query results page (see figure 4) will be displayed. Note, that queries may take up to five
minutes under heavy server load conditions. Furthermore, queries involving a lot of context data
and queries for aggregated data take longer to execute than queries for raw data and with few
context data.

Constraining queries to a user specific species list
Often, information is needed for a relatively small number of species only. In order to accommodate
this need, users may upload a list of species to the web query application and then use this list to
constrain query results.
      Figure 4: The query results page. Click on the encircled CSV file button to get the results sent via e-mail.

Step 1: Composing your species list
With a simple text editor (e.g. Editor, accessible from the start->programs->accessories menu of
the Microsoft Windows operating system), compose your species list by entering one species name
per line and save the resulting text file somewhere on your computer. Note, that webquery can only
interpret text files. If you use a word processor as e.g. Open Office or a spreadsheet program as e.g.
Microsoft Excel be sure to save your species list as plain text file.

Step 2: Uploading your species list
Click on the Species Selection tab at the top of the web query window. This will open the Species
Selection Window shown in figure 5.

                                         Figure 5: Species Selection Window
In order to upload your list, first select a species list against which the names in your file shall be
matched in the Select reference species list dropdown. Only matching names will be uploaded. The
lists available in the LEDA Traitbase are given in the appendix of this document. After this, you can
upload the species list file created in step 1 by clicking the Choose button in the right middle section
of the Species Selection View, selecting the appropriate file in the file selection dialog that pops up,
and clicking the Add LEDA species from file button.

Step 3: Organising the uploaded species information
The Species Selection View displays species name information in two tables. The names uploaded
from files appear in the lower table denoted as List of available but not selected species in figure 5.
You can consider this list as your private species name scratchpad in LEDA web query. Whenever
you load species names from a file they appear in this list unless they cannot be matched to valid
LEDA species names. Note, that names from previous uploads stay in this table unless they are
moved or deleted. Thus, species name uploads do not have to be repeated for each query. However,
since the list is not currently sorted (this is planned for a forthcoming version of webquery) and at
most three entries are displayed at a time, we decourage you from keeping too many entries in the
For uploaded species names to be used in queries, they have to be moved to the Currently selected
species table at the top of the Species Selection View. This can either be done by clicking the select
button figuring beside each species name in the lower table or by clicking the Add all species to
selection button at the lower right corner of that table. In the latter case all species names contained
in the lower table will be moved to the upper table.
By clicking the Remove button figuring besides each name in the upper table, you can move the
corresponding name back down to the List of available but not selected species names.

Step 4: Linking a query to the list of selected species names
Once you composed a list of species for use in a query, you have to bind the list to the query. In
order to do this, click on the Query Composition tab at the top of the web query screen. This will
open the Query Composition View. Assuming that you have already composed a query or selected a
saved query for editing, you can now click the Refine Query button positioned below the Sample
Query header display. This will open the Define Conditions view (see figure 6).

                        Figure 6: Define Conditions view without any defined conditions
Clicking the Add button will add a new condition line to the View (see figure 7).

                        Figure 7: Define Conditions View with a single condition entry
Now select the entry SBS number from the attribute selection dropdown at the condition line's left
side. This will automatically fill default values into the remaining two fields of the condition line,
resulting in the condition "SBS number IN User selected SBS numbers".
Clicking the Confirm and back to query button will activate the Condition. All queries will now be
constrained to the list of species names figuring in the Currently selected species list from the
Species Selection View.
Important Note: Until removing conditions by clicking the Remove all conditions button and then
the Confirm and back to query button in the Define Conditions view, they apply to all queries. So
be sure to remove the condition linking the species selection to your queries, if you want to execute
an unconstrained query.

Handling query results
Web query will limit the results displayed on the browser in order to keep memory requirements on
your system low and to reduce network bandwith usage. If the displayed result set is limited, LEDA
web query will warn you.
If you want the query result sent as comma separated value file attached to an e-mail message, you
may notify web query to do this by clicking the CSV file button above the list of query results. Note,
that the number of records sent via e-mail is limited for guest users. Normally, you can work around
this by precisely formulating your queries using the refine queries feature (not documented in the
current version of this guide). Please contact us at if you want to analyse
bigger amounts of data for scientific purposes. We will be happy to cooperate.

Common problems with query results
Often, only a table header is displayed as result of a query. This is caused by the query being
Should you encounter this problem, first look up, whether the null values allowed box has been
checked in the query compostion view. If you did not check it, try to do that and execute the query a
second time by clicking the execute query button. If the box is not checked only rows of data
having values for each of their cells will be displayed in the query results. For example, if a you
want LEDA web query to display raw data records with minimum, maximum, and average values,
but the Traitbase raw data currently does not contain minimum values for the given trait, no records
will be displayed, should the null values allowed option not be given.
If you checked the null values about box, but the query result still does not contain any records,
check whether you defined conditions with the refine query option accessible from the query
composition view. Note, that such conditions are currently not removed automatically in the current
version of LEDA web query, even if you start to compose new queries. To remove these "dangling
conditions" click the refine query button in the query composition view, which opens the refine
query view. At the bottom of this view, first click the remove all conditions button and then the
confirm and back to query button.
An other problem that may occur due to the early version of the LEDA web query tool results in an
error message being displayed instead of a query result. Should this occur, please send an e-mail to containing a list of the attributes your query is composed of together
with the approximate time of execution. We will then use your information for improving the
quality of the LEDA web query tool.


Trait attribute descriptions
For technical reasons, the trait attribute identifiers appearing in the select trait attributes drop down
and in query result headers are limited to a maximum length of 30 characters. Thereby, they are not
very self-explanatory. While a forthcoming version of the LEDA query tool will contain an online
help for describing the trait attribute identifiers meanings, you currently have to rely on this
section's information and on the LEDA data standards for attributes at the raw data level.
Attribute identifiers may consist of three parts. First, an identifier may be prefixed by an optional
mnemonic describing the nature of the attribute. The meaning of these mnemonics are given in table
1. The second part of an attribute identifier is a short description of the attribute's meaning. It often
includes a unit designator. If this identifier part contains the string single value, the attribute is
calculated from multiple entries in a trait record and is used to have a basis for aggregation
operations even in the case of sparse data.
Mnemonic Meaning

ADM           Administrative information – normally not visible

AOF           Trait specific attribute for Age of first flowering

BBS           Trait specific attribute for Bud bank - seasonality

BBV           Trait specific attribute for Bud bank – vertical distribution

BRA           Trait specific attribute for Branching

BYC           Trait specific attribute for Buoyancy

CGO           Trait specific attribute for Clonal growth organs and Role of clonal growth organs

CH            Trait specific attribute for Canopy height

DIA           Helper trait specific attribute for Diaspore morphology

DIS           Trait specific attribute for Dispersal

END           Trait specific attribute for Internal animal dispersal

EPI           Trait specific attribute for External animal dispersal

GEO           Context attribute carrying Georeference information

HAB           Context attribute carrying Habitat type information

HC            Trait specific attribute for Habitat characteristics

LAT           Trait specific attribute for Lateral spread

LD            Trait specific attribute for Leaf distribution along the stem

LDC           Trait specific attribute for Leaf dry matter content

LF            Trait specific attribute for Leaf form

LM            Trait specific attribute for Leaf mass

LS            Trait specific attribute for Leaf size

M             Context attribute carrying Collection and measurement Method information

MDU           Trait specific attribute for Morphology of dispersal unit

NAM           Context attribute carrying Naming information
Mnemonic Meaning

ORG           Context attribute carrying information on Original reference

PER           Trait specific attribute for Persistence of mother-daughter connection

PGF           Trait specific attribute for Life form

PLS           Trait specific attribute for Plant life span

PLT           Context attribute carrying information on Plant stage

PSS           Trait specific attribute for Period of seed shedding

REF           Context attribute carrying information on Reference

RH            Trait specific attribute for Seed releasing height

SCF           Trait specific attribute for Seed crop frequency

SF            Trait specific attribute for Shoot form and associated Branching

SLA           Trait specific attribute for Specific leaf area

SM            Trait specific attribute for Seed mass

SNP           Trait specific attribute for Seed number production

SS            Trait specific attribute for Seed shape

SSB           Trait specific attribute for Soil seed bank

SSD           Trait specific attribute for Stem specific density

STM           Context attribute carrying information on Support structure

TV            Trait specific attribute for Terminal velocity

Single values for numeric traits are most often calculated using the following code:
        IF median value available from record
             THEN single value := median value
        ELSE IF mean value available from record
             THEN single value := mean value
        ELSE IF minimum value and maximum values are both available from record
             THEN single value := (maximum value + minimum value) / 2
        ELSE IF minimum value available from record
             THEN single value := minimum value
        ELSE IF maximum value available from record
             THEN single value := maximum value
        ELSE single value := NULL

Whenever the second part of an identifier contains the string (sv) it means, that the attribute value's
calculation is based upon the single value attribute.
The third (optional part) of an attribute identifier is the appendix (aggr), denoting that the attribute's
value is calculated using aggregation operators as e.g. mean, minimum, modal value but also
aggregation operators specially defined for the LEDA Traitbase as e.g. index operators. Attributes
whose identifiers end with (aggr) may only be used in aggregated queries. In that case the
remaining attributes also figuring in the query attribute list are used as means of grouping records to
be aggregated.

Species lists currently available in the LEDA Traitbase
The following table maps the entries from the species list dropdown in the Species Selection View
to species lists.
Name in species list dropdown                       Species list reference

STUK species list                                   Stace UK species list

NO species list                                     Norway species list

FI species list                                     Finland species list

BE species list                                     Belgium species list

UK species list                                     United Kingdom species list

SCAN species list                                   Scandinavia species list

D species list                                      Häupler & Wisskirchen species list

CEN_EU species list                                 Central Europe species list

P species list                                      Poland species list

F species list                                      France species list

LV species list                                     Latvia species list

LT species list                                     Lithuania species list

LEDA species list with numbers from SBS             Synbiologisch Taxonomisch Systeem species
(edition summer, 2003)                              list with long species names (old version)

SBS species list                                    Synbiologisch Taxonomisch Systeem species
                                                    list with long species names (current version)

NL species list                                     The Netherlands species list

Synonyms from BIOPOP

SBSshort species list                               Synbiologisch Taxonomisch Systeem species
                                                    list with short species names (no authors)

NINA species list

Shared By: