Supplementary information for CCH search and detail pages
Most of the herbaria have a number fixed to each specimen. This number, sometimes called the accession number, serves as a unique identifier in the database when associated with a herbarium abbreviation. Some herbaria have such a number only for the database, and it does not appear on the sheet.
The name under which the specimen is currently filed. Note that participating herbaria may use different names for what is the same taxon. The author of the name is not given, but may be retrieved, along with other taxonomic information, by clicking on "More information: Jepson Online Interchange".
The collector is the person or people recorded on the label as having collected the specimen. Only the last name of the first collector is used for searching and sorting. That is, Hall and Michener will sort with H. M. Hall and a search for Michener will not turn up specimens collected by Hall and Michener. The collector name is linked to the Harvard Index of Botanists, or for a few collectors, to other sources of biographical information. The collector number is the number used by the collector to index the specimen. Collector numbers have a variety of formats. For the purposes of searching and sorting, the collector number is assumed to contain an alphanumeric prefix followed by a numeric stem followed by an alphanumeric suffix, of which only the stem is pertinent. The date is the date or date range on which the specimen was collected. It is stored in the database both as it occurs on the label, and as early and late Julian dates. The Julian dates, which are used for searches, accommodate date ranges. Searches for specific dates (e.g., June 6 1946), will also retrieve specimen data with a date range that includes the specific date (e.g., Jun 1946 or 1946). The date on the detail display is formatted as in the database. The date in the primary result set is formatted consistently: month, day, year.
Jepson's collection numbers are linked to images of his field books, as are Brewer's and Bolander's Geological Survey collection numbers, and some of the numbers of C.A. Purpus and Annie Alexander.
The California county (spelling standardized) as it appears on the label, unless that is in error. Note that during the history of botanical collection in California, county boundaries have been altered several times. If the county does not appear on the label it has either been entered as "Unknown" or secondarily determined by a data enterer or user.
The name of the place the specimen was collected, or a description of how to reach that place. Locality information is not standardized in any way and a wide variety of spellings or usages for the same place may be encountered. Users should not expect to retrieve the data for all specimens collected from Yosemite National Park by searching for "Yosemite". The SMASCH database at Berkeley was designed with four locality fields: loc_place, for a referenced place (e.g., Clark's Ranch); loc_distance, for an indication of direction or distance from the referenced place (e.g., 3 mi NW of); loc_other, for other locality terms on the label (e.g., Yosemite --usually more encompassing terms); loc_verbatim (a more or less verbatim capture of the label data, usually used in place of the other three fields). In the Consortium display, these four fields are concatenated, with
an attempt made to approximate the original label and reduce redundancy. The attempt is algorithmic, and not 100% successful.
Locality searches: Words in the locality search box are
anded. Results will be returned only if all the words are found. Very common words
are disregarded. If the search box is filled with a quoted 2 or 3 word phrase,
only records containing that exact phrase will be returned.
Geographic region searches: Geographic regions are defined by rectangular bounding boxes; therefore, it is likely that some anomalous records will be included. Furthermore, records can only be returned if they include coordinates, so some will be missed.
If you want to search within an elevation range, fill out both the Upper and the Lower field.
On the detail page the elevation in feet or meters (units indicated) of the collection site. Elevation may be determined by instrument, map, or estimation. Accuracy should not be relied upon. Apparent accuracy may result from conversion. For the primary result set, all elevations are displayed as meters.
Physiographic or biological details about the collection site.
Here is a brief account of how longitude and latitude coordinates get into
our database, which only includes California localities:
From coordinates on the label or in a database (less than 10% of specimens).
Source and precision is not known. Errors on
the label are frequent and derive from misreading maps, mistyping figures,
missetting GPS units, and computer operation errors. Labels are often not
made by the collector, but are transcribed from a fieldbook or other
medium by someone who may or may not be familiar with the collection
region. The label typer sometimes adds coordinates
From descriptive information on the label, evaluated using archival material
at the herbaria: collector's field books/itineraries, historical maps, etc.
Converted from TRS data on label.
Many labels bear Township, Range, and Section data without other
coordinate data. These figures can be converted using the program TRS2LL.
If no section is given, we assign a section in the middle. Coordinates
assigned this way can be off by several miles (or much more if the original TRS numbers are not
accurately entered). We have found that TRS coordinates are troublesome because
is easy to misenter them originally from a map, and easy to confuse east and
west, north and south, or mistype the numbers. There is the additional
complication of 3 baselines in California.
Assigned by the Biogeomancer program. Biogeomancer can automatically
parse an English locality description and assign latitude and longitude.
Unfortunately, it is (in its current incarnation) usually unhelpful
because of its assumptions.
For example, 4 miles W of Coalinga is placed 4 air miles due west of
Coalinga, whereas the location is probably 4 road miles along a road
trending west rather than east. The location of a stream is given as the
mouth of the stream. We use Biogeomancer to supply coordinates for
localities that are represented as one place--Coalinga, Peanut, Yreka. We
don't currently supply an error for Biogeomancer-derived coordinates. The
Biogeomancer program is being completely remade currently and should
eventually be much smarter.
Assigned by people interpreting the databased locality data using the
mapping program Terrain Navigator, which allows quick lookup of a
referenced locality and fairly easy measuring along a road or line. We
extract locality data into an Access file by county and one georeferencer
works on one county. The error assigned in this process takes into account
uncertainty of the reference point and other uncertainty. This system was
developed by the Museum of Vertebrate Zoology at Berkeley.
Regarding herbarium specimens from cultivation: In these cases there are two
localities of interest: the place where the plant was cultivated, and the
provenance of the source material: the place in the wild from which seeds,
bulbs, rhizomes, cuttings, etc. were derived. Both localities are useful and
interesting in different ways, but only one set of coordinates can be mapped.
In an effort to most accurately represent the biodiversity of California plants,
the CCH prefers to display coordinates for source material, if that
information is available. When only the locality of cultivation is available,
we prefer to still serve the record while masking the coordinates; in some
records, however, the fact that a plant was cultivated is not expressed precisely
enough that it can be automatically detected.
When specimens are mapped to their garden location, they are very likely
to display as a yellow flag. In these cases we encourage users to write to
David Baxter so that the record can be modified accordingly.
The BerkeleyMapper application plots position using the coordinates from the database; it does not assign coordinates.
NOTE: Coordinates have not yet (Winter, 2012) been systematically verified.
Derivation of latitude and longitude figures. Herbaria participating in the Consortium use a variety of protocols for
evaluating locality descriptions and obtaining coordinates.
coordinates for specimens that lack a longitude-latitude statement are found using
MAPTECH Terrain Navigator, backed up by SBBG map archives and field notes
(especially for Channel Islands and military bases). As of Winter, 2006, coordinates are not uniformly tagged
as to their derivation. If the source is listed as "Not recorded", it could have come from a label or it could have been derived secondarily. If the coordinates are obtained by georeferencing at UC/JEPS using the Terrain Navigator, Biogeomancer, or TRS2LL programs, then this is indicated.
Township, Range, Section are frequently indicated on labels, and not always correctly. It is easy to mistranscribe TRS from a map and easy to miscopy the combination of numbers and letters. In California, there is the additional complication of three meridians: Humboldt, Mount Diablo, and San Bernardino.
...Read Wikipedia article
Information from a variety of fields. Sometimes merely the fact that a certain kind of information occurs on the label is recorded.
Some of the participant databases store annotations that have been posted on the specimen sheet: reidentifications, assignments to an infraspecific rank, etc. Sometimes, in accordance with a recent taxonomic treatment, a curator transfers all specimens assigned to a species from one genus to another, without evaluating the specimens. This is what is meant by "curatorial action". Usually, but not always, the specimen is filed in accordance with the most recent annotation.
We welcome comments and corrections of the specimen data that we display here. Action on the comments will be taken by the curators of the participating herbaria. You should provide helpful explanatory information if you can. Incorporating the comments into the database is time-consuming because in most cases it is necessary to examine the specimen label before amending the database. Often a discrepancy can be caused by more than one error, and investigation of an error sometimes reveals related problems.
These are searches that users (especially curatorial users cleaning up data) are likely to want to run next. For example, from this search:
- taxon=Quercus wislizeni
- county=Solano; taxon=Quercus wislizeni
- county=Solano; collector=Jepson
- collector=Jepson; taxon=Quercus wislizeni
- collector=Jepson; number=2378
- collector=Jepson; number=2378; include nearby numbers
- date=Oct 18 1903; collector=Jepson
- date=Oct 18 1903; county=Solano
- Return geographically nearby records (within ~3 mi).
List of names, 1 record each
- Plot elevation range (from Jepson Manual) against latitude
- Possible GBIF records [External]