Wildl. Soc. Bull. 22:126-128, 1994

                            WILDLIFE SOFTWARE
                                    
            HEICALC: ELK HABITAT-EFFECTIVENESS INDEX SOFTWARE

ALAN AGER, U.S. Department of Agriculture Forest Service, Pendleton, OR 97801

MARK HITCHCOCK, 2347 Old Day Creek Road, Sedro Woolley, WA 98284

Key words: cover, elk, forage quality, forage quantity, habitat-effectiveness index,
habitat models. roads


     Thomas et al. (1988) developed a Habitat Effectiveness Index (HEI) to measure elk
(Cervus elaphus) habitat quality for winter ranges in the Blue Mountains of eastern Oregon.
This model evaluates (1) the size and spacing of cover and forage areas, (2) the density
of roads traveled by motorized vehicles, (3) the quantity and quality of forage, and (4) the
quality of cover. The four habitat components are quantified and combined in an overall
HEI as follows:

HEIsrfc = (HEs x HEr x HEf x HEc)^1/4

where HEs = habitat-effectiveness index derived from the size and spacing of cover and
forage areas, HEr = habitat-effectiveness index derived from the density of roads open to
vehicular traffic, HEf = habitat-effectiveness index derived from the quantity and quality of
forage, and HEc = habitat-effectiveness index derived from cover quality.

     This model has been incorporated into federal and state land management plans
to monitor elk habitat. Unfortunately, its application has been constrained by the time and
effort required to manually calculate the subindex that describes the spatial arrangement
of forage and cover areas. An earlier effort to automate HEI calculations (Leckenby et
al.1985) produced a mainframe program that suffered from limitations commonly
associated with mainframe software (e.g., accessibility, cost, data input and management).
Abbreviated procedures for calculating the subindex have also been developed; however,
their use has led to a lack of standardized calculations (Edge et al. 1990). HEI calculations
have also been programmed into geographic information system (GIS) software, but this
technology is not readily available to many state and federal wildlife biologists in the Pacific
Northwest.

     Our solution was to develop simple menu driven DOS microcomputer software for
the HEI model (Hitchcock and Ager 1992). HEICALC calculates HEI's for the Blue
Mountain winter range model from a digital habitat map and data on roads and forage.
Software for the western Oregon model (Wisdom et al. 1986) also has been developed and
is described elsewhere (Ager and Hitchcock 1992). In addition to calculating HEI's,
HEICALC can be used for other landscape analyses where habitat must be classified
according to its distance from edge (e.g., old-growth fragmentation).


                               HEI SOFTWARE
                                     

     HEICALC generates HEI's according to the Blue Mountain winter range model
(Thomas et al. 1988) using ASCII habitat maps and data on forage and open road
mileage. The habitat map depicts cover and forage with integer values that represent
pixels of a fixed area. The pixel size can be varied by the user, although for HEICALC
to function as intended pixels should have dimensions evenly divisible into the 300-foot
(91-m) distance bands in the HEI model. We commonly use pixels of 150 feet (46 m) on
a side, which encompasses 0.52 acre (0.21 ha). Our tests with cover maps of winter
range in the Blue Mountains showed that larger pixel sizes resulted in a loss of small
"stringers" of cover during the raster conversion of polygon maps. These small cover
areas may not be significant in terms of elk habitat use (Thomas et al. 1988), but their
deletion from the habitat map should be performed by wildlife biologists rather than by
GIS software. In addition, with pixels larger than 300 feet (91 m) on a side, the diagonal
distance between pixel centers exceeds 1 distance band, thus all habitat comparisons
between diagonal pixels contribute to the 301- to 600-foot (92- to 183 m) spacing band.
This problem is eliminated with the smaller pixel sizes. The only disadvantage to using
a smaller pixel size is that processing time increases as the map arrays become large.

     Habitat code data are entered interactively or read from ASCII files generated
with raster based GIS. Map data are displayed in a spread sheet along with the
coordinates and habitat type of the cursor location. The cursor location on the habitat
map is controlled via keyboard arrow keys or a mouse. The habitat code at the cursor
location is edited by entering a habitat code from the keyboard. Habitat maps can be
printed on most printers using standard ASCII printer characters to control formatting.

     Output from the program includes detailed classification of habitat types by
distance bands. Each of the 4 HE component values are displayed along with the
resultant cumulative HEI with and without the forage HE score. Total edge length of
cover areas also is included. Output screens can be printed or sent to a file. A
descriptive title can be added to the output if desired. A data base feature allows the
user to link maps to an ASCII data base containing up to 5 attributes. Attribute data can
be retrieved when viewing the map.

     HEICALC can be used for any habitat buffering problems by creating HEICALC
maps coded to reflect the resource of interest. For instance, we use HEICALC to
measure acres of "interior" old-growth patches by coding old growth as cover and the
remaining area as forage in the habitat map.

     HEICALC was written and compiled using Borland International TURBO
PASCAL1 7.0 compiler and the 8087 emulator library (1800 Green Hills Road, Scotts
Valley, CA 950670001). The program is compiled in DOS real (vs. protected) mode and
will run on IBM-PC compatible computers. No installation routine is necessary.
HEICALC automatically detects the video display type and takes advantage of display
capabilities. The program uses a dynamic memory allocation scheme that uses nearly
all available standard memory. Memory requirements depend on project area. At least
130 kilobytes of available memory is required to load the program and process habitat
maps of 60,000 cells. The maximum map size accommodated by the program is 600
columns by 800 rows, which requires 540 kilobytes available memory.

     Acknowledgments.--Partial funding for this project was provided by the Blue
Mountain Elk Initiative.

' The mention of commercial products does not constitute endorsement by the U.S.
Department of Agriculture.

                             LITERATURE CITED
                                     
                                     
AGER, A. A., AND M. E. HITCHCOCK. 1992. Microcomputer software for calculating the
Western Oregon elk habitat effectiveness index. U.S. For. Serv. Gen. Tech. Rep.
PNW-GTR-303. 12pp.

EDGE, W. D., S. L. OLSON-EDGE, AND L. L. IRWIN. 1990. Planning for wildlife in
National Forests: elk and mule deer habitats as an example. Wildl. Soc. Bull. 18:87-98.

HITCHCOCK, M. E., AND A. A. AGER. 1992. Microcomputer software for calculating elk
habitat effectiveness index on Blue Mountain winter ranges. U.S. For. Serv. Gen. Tech
Rep. PNW-GTR-301. 12pp.

LECKENBY, D. A., D. L. ISAACSON, AND S. R. THOMAS. 1985. Landsat application to
elk habitat management in Northeast Oregon. Wildl. Soc. Bull. 13:130-134.

THOMAS, J. W., D. A. LECKENBY, M- HENJUM, R. PEDERSEN, AND L. D. BRYANT.
1988. Habitat effectiveness index for elk on Blue Mountain winter ranges U.S. For. Serv.
Gen. Tech. Rep. PNW-GTR-218. 28pp.

WISDOM, M. J., L. R. BRIGHT, C. G. CAREY, W. W. HINES, R. PEDERSEN, D. A.
SMITHEY, J. W. THOMAS AND G. W. WITMER. 1986. A model to evaluate elk habitat in
western Oregon. U S. For. Serv. Publ. R6-F&WL-216-1986. 27pp.


Received 20 April 1992. Accepted 13 September 1993. Software Editor: Rexstad.