Grass Manual: Notes on Distributional
The grass distribution maps posted at this Web site are generated
by ArcView from the geographic database developed by Barkworth
for the Manual of North American Grasses. They are
based on a wide range from sources (Barkworth
and Capels 2000; current
Each map draws on at least two, and possibly four, data files.
One file contains county level data for the contiguous 48 states
of the U.S.A. The color of the shading indicates the source
of the record.
The other three files are all locality files, meaning that
they contain latitude and longitude data, plus information about
the data source. One of these contains data from diverse sources,
ranging from specimens to publications of various kinds. The
second locality file is generated from the Intermountain Herbarium
database; the third locality file is similar in structure to
the Intermountain Herbarium database but contains information
from other herbaria or from individual collectors.
We have only just begun to develop the "Other Herbaria"
file, so it does not contain many records. Our ultimate goal
is to have the maps drawn by calling on information in the databases
of individual herbaria.
At present, the maps are prepared by generating jpg images
and posting these images to the web. We are working on making
them available dynamically and providing active links from the
locality dots to the files from which they are generated so
that anyone interested can find out the basis of the record.
Sources of Data
Some contributors provided us with county level information.
Such records appear in red. Others provided shaded area maps
which we could not easily enter into the database so it appears
as if we have no data from the contributor, but this is not
necessarily the case. If we were given locality data, the data
will show as a blue dot on the maps and the information provided
is in the database.
Some contributors were not able to provide us with distribution
information because they are not employed by institutions or
companies that can provide the resources needed for such work.
It should be noted that contributors received no support for
preparing their treatments from the project. Some have positions
that encourage such work; many do not. Without the generosity
of the contributors, projects such as the Manual would
sources of data included atlases, checklists, and a variety
of other publications; databases from herbaria or, in the case
of weeds, agencies within the USDA; and from a variety of individuals
who were so kind as to send us their data.
The areas shown in color correspond to counties. Most
herbarium specimens from the contiguous United States
note the county in which a specimen was
collected, and counties are frequently the unit of record
for state distribution maps and for checklists. This
makes it relatively easy to acquire county level data.
When a county has been shaded it merely means that the
species concerned has been found at least once in the
county, not that the species occurs throughout the county.
The colors indicate the class of data on which a county record
The only verifiable distribution maps are those based on herbarium
specimens. Ideally, all the maps in the Manual would
be based on specimens that had been identified by the contributors.
For most genera, that is impossibly idealistic. There are simply
too many specimens in too many herbaria for any individual to
check, particularly individuals that are receiving no compensation
and no staff support for their work, the condition of many contributors.
A second best approach is to ask taxonomists in various
parts of the country to check the specimens in their
region, using the treatment developed by the contributor.
This too is an enormous task, and
one that has not been considered in the Manual's funding
to date. We are soliciting funding to fill in some of
the regions for which we have few data but we cannot,
realistically, expect to obtain the funding needed to
complete all the maps in this manner.
Our database takes an approach that many would not consider
acceptable: we enter published data from a variety of
sources, often not knowing who has done the identification
nor what treatment they followed. If the Manual contributor
adopts a narrower interpretation of a species than the
person whose work we used in compiling the database,
our maps will be seriously in error. Fortunately, the
number of times that this has happened is relatively
few - but it is a problem with some species.
We shall send all maps to the contributor before publication,
even publication on the Web, but this may not resolve
all the problems. In many respects, preparation of detailed
maps should be delayed until after treatments have been
published. We prefer to provide maps and encourage those
who find them in error, or questionable, to draw attention
to the specific problems. We shall create a message
center where such questions can be posted and others,
as well as the contributors and ourselves, can respond.
Since "ourselves" currently consists only
of Kathleen Capels and myself, it is unlikely that
we shall be able to follow up on any but the major problems.
The very best solution would be to have a Web site that would
provide dynamic distribution maps in which each locality record
is linked to a herbarium record. Technologically speaking, it
is very doable - but no one wants to pay for the unglamorous
and time consuming task of placing herbarium data into a database.
Many herbaria have begun to do so, so a pilot scheme in which
their data are made available is feasible. Once available, anyone
who needed to be sure that a particular distribution record
was correct would be able to determine where the documenting
specimen was located and inquire about having it verified, in
return, of course, for appropriate compensation for the work