He, 1993). Vegetation coverage, a GIS term,
was used to distinguish this work from
traditional plant inventories and vegetation
mapping. The inventories do not attempt to
create complete and continuous data such as a
map or coverage about species composition of
floristic communities. Traditional hard copy
vegetation maps are fixed in time and
resolution, unlike a GIS coverage, which can
incorporate new data as acquired. The
vegetation coverage employed here, represents a
GIS approach towards an integrated
manipulation of floral data. It has the
characteristics listed below.
* GIS digital vegetation layers, which
contain polygons of species information
and associated database records. The
database contains multiple attributes of
each polygon, such as dominant tree
species, co-dominant tree species,
identifications in a multi-level forest
classification system, etc.
* Several other spatial data layers which
influence the vegetation or the
interpretation of the coverage, such as
topography, soils, hydrography, and
administrative boundaries.
* A flora database from field plots, which
incorporates all available field inventory
records. Precise plot locations are
determined and linked with the above
databases. The flora database can be
used to support many spatial and non-
spatial queries and analyses, such as
species distribution or spatially
associated species analysis.
+ A flexible, hierarchical vegetation
classification based on species
composition.
With the above characteristics, the vegetation
coverage can provide hard copy vegetation maps
at different scales and classifications, and also
can support complicated spatial analyses which
are difficult by traditional methods, such as
selection and display of critical tree species
distribution (e.g., hemlock, white cedar), and
analysis of site conditions on which these occur.
60
METHODS
The project integrated previous studies with new
data from a recent flora inventory (Judziewicz
and Koch, 1992), air photo interpretation and
other environmental GIS data layers (Ventura
and He, 1993).
Data from Previous Vegetation Inventory
Data from a recent vegetation inventory
conducted by Judziewicz and Koch during 1991
and 1992 were compiled by National Park
Service staff into a flora database (FLORADB).
Flora were observed on a regular grid of 200,
300, or 400 meter spacing depending on the size
of the (Eagle, Gull, Long and the Mainland Unit
were not included) (Judziewicz and Koch,
1992). These plots were quite systematic and
complete in terms of both plot location and plot
attribute recording. The database incorporated
1424 field plots. It was converted to INFO
format with more than 20,000 records, each
representing a species description. The
information derived from FLORADB can be
used for species based analyses, such as
dominance, commonness, and so forth.
Since each surveyed plot had a recorded
location, we built those plots into a point
coverage. The point coverage can be used in
association with FLORADB for many on- screen
queries, or overlaid with other data layers such
as a data layer from air photo delineation to help
interpretation.
Integrated Terrain Units (ITUs)
ITU is a coverage which incorporates several
environmental factors. It is typically generated
from physical and biotic factors influencing
vegetation, such as slope, aspect, soils,
hydrography, and etc. Thus each ITU polygon
has a unique combination of slope, aspect, soil
type, and hydrological conditions. For the
Apostles, these data were automated only where
source material of sufficient quality to contribute
to vegetation characterization was available
(hydrography and terrain were used on all
islands; soils were less available).
All the data layers were prepared in the Arc/Info
format from existing paper or mylar maps.
These I
covera£
Since t!
determi
photogi
were er
vegetat
with a
vegetat
NN
Veget
Species
forest st
classific
Island
and con
overstoi
hierarcl
southea
in most