Computer-Assisted Habitat Mapping for Black Bear Management in Shenandoah National Park

COMPUTER-ASSISTED HABITAT MAPPING FOR BLACK BEAR MANAGEMENT IN SHENANDOAH NATIONAL PARK

JAMESF.WILLIAMSON, Jr.,1 Departmentof Fisheries and Wildlife Sciences, VirginiaPolytechnic Institute and State University, Blacksburg, VA 24061 JAMESB. WHELAN,2Virginia Cooperative Wildlife Research Unit, VirginiaPolytechnic Institute and State University,Blacksburg, VA 24061

Abstract: The objective of black bear (Ursus americanus) management in Shenandoah National Park has been to minimize propertydamage and personalinjury to Park visitors while maintainingthe bear populationas a partof the naturalfauna. Past management attention has been directed at educating visitors; however, future efforts will incorporate more biologically oriented strategies, and will require new information on ecological matters such as bear habitat suitability, the location of areas of sensitive or critical habitat, and the impacts of proposed developments on bear habitat. A computer mapping system designed to as- sist in the bear management effort by meeting these types of information needs is described. Sample maps were prepared.

lnt. Conf. Bear Res. and Manage. 5:302-306

The 483-km2Shenandoah National Park lies in at a level compatible with the habitat. To effect a narrowand irregularstrip along the crest of the this end, a sociological strategy aimed at educat- Blue Ridge Mountains from Front Royal, Virgin- ing Park visitors was adopted. In 1975, a public ia, southwest to Waynesboro, Virginia. A park- relations campaignwas launched in which visitors way, the Skyline Drive, provides primaryvehicu- were warned of the physical hazards and legal lar access and bisects the Park along the major consequences of feeding Park wildlife. Baptiste mountain ridge. et al. (1979) found that this campaignwas gener- Although the area was logged repeatedly prior ally successful and that most visitors were aware to the establishment of Shenandoah National of the bear problem and the dangers involved in Park in 1934, almost all of the Park is now a feeding campgroundbears. Evidence for success closed-canopy forest. Soils are sandy and rocky, of the programwas also found in reductions in and upper slopes and ridges are dominated by xe- damage incidents from a high of 231 ($14,672 in rophytic oaks (Quercus spp.) and pines (Pinus damages) in 1975 to a low of 54 ($1,868) in 1979 spp.). Coves, toeslopes, and more gradual (Shenandoah Natl. Park, unpubl. rep., Luray, midslopes contain a variety of mesophytic species Va., 1978; W.E. Phillips, ShenandoahNatl. Park, typical of the northern Blue Ridge (Braun pers. commun.). 1950:221-225). A biologicallyoriented strategyinvolving more Shenandoah provides habitat for a variety of direct management of the bear population was wildlife species, including both small and larger developed to complement the visitor-orientedef- mammals, notably white-tailed deer (Odocoileus fort. In an attempt to encourage bears to rely on virginianus)and black bear. Recent bear popula- natural foods, Park landfills were closed in 1975, tion estimates range from about 150 animals (Du and campgroundand picnic area garbagerecepta- Brock 1980) to approximately300 (L.L.Hakel, cles were made more secure against rummaging unpubl. rep., ShenandoahNatl. Park, Luray,Va., bears. Persistentlyobnoxious bears were trapped 1977). and relocated in remote areas both inside and As in most national parksin the United States, outside the Park. a primaryobjective in the Shenandoahbear man- In 1972 the VirginiaCommission of Game and agement plan was to minimize propertydamage Inland Fisheries, in conjunctionwith Shenandoah and personal injury resulting from encounters National Park, initiated researchdesigned to esti- with bears, while maintainingthe bear population mate the bear populationsize and to gain insight into the bear population structure (Raybourne 1 Present address: Division of Land and Forest Resources, 1976). Basic biologicalinformation of this nature Tennessee Valley Authority, Norris, TN 37828. has begun to provide insight into bear ecology in the and ef- 2 Present address: U.S. Fish and Wildlife Service, 2629 Shenandoah. However, development Redwing Road, Fort Collins, CO 80526. fectiveness of a comprehensive, long-range bear

302 COMPUTER-ASSISTEDHABITAT MAPPING * Williamsonand Wlhelan 303 plan for Shenandoah now depends on the avail- Table 1. Importance values and intensity levels for 4 criteria used to evaluate black bear habitat in Shenandoah National ability of additional information on bear ecology Park, Virginia. in the Park. Questions concerning bear habitat suitability,the location of sensitive habitat, carry- Relative Relative intensity importance ing capacities, saturation densities, movements, Habitatcriterion value value and the impacts of natural- and man-induced 1. Proximityto potential 3 poachingsource changes in bear habitathave yet to be addressed. a) outside Park 0 The purpose of this research was to develop a b) within 1.6 km of 2 peripheralaccess to Park methodology for producinga few select types of c) over 1.6 km from 9 informationrelevant to the managementof bears peripheralaccess to Park in the ShenandoahNational Park. 2. Distanceto campgroundor 9 picnicarea We gratefully acknowledge the Shenandoah a) over 1.6 km 0 National Park who offered informa- b) within 1.6 km 1 personnel 3. Distancefrom regularly 8 tion and assistance during the progress of this travelledroad or development thanks are due Park a) within 1.6 km 0 study. Special Biologist b) over 1.6 km 9 W.E. Phillips, Chief Park Ranger Larry Hakel, 4. Forest type preferenceindexa 9 and Superintendent Robert Jacobsen for their a) closed oak type 9 b) open oak type 3.4 cooperation. c) cove hardwoodtype 3.7 a Decimal values for relative intensity were obtained by standardizing METHODS Beeman's(1975:94) preference to a scale of 0-9. The product of this research was MAP4B, a computer mappingsystem designed to manipulate tive importanceand intensity values to respective spatial habitat data and to display results in map criteriawere based on literature,opinions of vari- format. The MAP4B system was composed of 3 ous black bear researchers,and opinions of Park major functional parts: (1) a habitat assessment personnel familiar with the distribution, move- component consisting of a series of programsfor ments, and relative density of bears within the reading in and manipulating raw habitat data, Park. Although these values were assigned sub- (2) a display program for producing line printer jectively, they represented the best available in- maps, and (3) a series of special-purposeauxilia- formationspecific to bear ecology in Shenandoah. A linear ry programs for changing or manipulating map additive model was used in the assess- data. ment algorithm. For any given location (cell), a The habitat assessment procedure utilized 4 qualitative measure of habitat suitabilitywas cal- key habitat criteria: (1) proximity to potential culated using the formula: poachingsources, (2) distance to potential unnat- n ural food sources such as V= b, x,j (1) campgroundsand picnic i-I areas, (3) distance from human disturbance,and (4) a black bear forest type preference index de- where veloped by Beeman (1975:94) that integratedfor- Vj = the habitat value of the jth location age and cover availability. A relative importance (cell), value indicating the importanceof a criterion to bi = the relative importance value of the habitat suitability was assigned to each of these ith habitatcriterion, criteria. These values ranged from 0 (not impor- xij = relative intensity value of the ith habi- tant) to 9 (extremely important). Similarly, 10 tat criterion, categories were established to accommodatevary- i = 1, 2, ..., n, ing amounts or relative intensity (e.g., distance to j = 1, 2, ...,m, a picnic area) of each criterion. Integer values m = number of locations, were assigned to categories in a linear manner n = number of habitatcriteria. an along interval scale on which the upper end Standard1:24,000 scale United States Geologi- correspondedto the ideal situation (Table 1). Se- cal Survey topographicmaps were used as base lection of habitat criteriaand assignment of rela- maps, and all map data were registered to base 304 COMPUTER-ASSISTEDHABITAT MAPPING * Williamsonand Whelan

l 2 34547@09 I234SX7&&$1 2 34S67999 1 234561."4 0 1 3 IS1.S31L 23]4S6*llb1 2 MG5790'1 l 45*7|||81 2 345s7|I1 I 34567|41 II4567|691 } 4S676i Modifications of 2 packaged programs (Federa- tion of Rocky Mountain States 1977) were used to convert recorded polygon data into proper format and to produce punched cards. The display feature of MAP4B was used to produce individual criteria maps. When these maps had been checked for coding errors, a final map of bear habitat suitability was produced with MAP4B using standard matrix addition to sum the weighted criteria maps (matrices). That is, each criterion map was weighted by a scalar (b,) and summed on a cell-by-cell basis to produce a matrix of sum values, [V]. The resulting cell 7; values in the composite map were stratified lin- early, and quantitative differences in habitat suit- i ability were displayed as 1 of 10 contrasting S symbols. I? Products were a series of computer line-printer maps measuring 30.5 cm by 38.1 cm. Row and I . I , column numbers were provided along the mar- gins to serve as numerical reference points. A scale and a legend showing frequency counts and value ranges by symbol were supplied with each ------..- -- 1I--..... -- --- map. ..r~E..... Special-purpose programs were developed to extend the utility of the basic MAP4B mapping :08. o.:,8 system. One such program was capable of pro- .1. 30,. O.-'. .. I4 I ducing interspersion maps based on the number

o .. o . , of different vegetation types present within a giv- [ill. - ]. 13S ' en radial distance of each cell in the map grid. I O . 0.0 Another auxiliary program was used to generate I ,8S l.'@ maps delineating zones or distances from fixed locations. In addition, programs for changing Fig. 1. Computer-generated map showing bear habitat suitabili- map display symbols and determining slope and Darker indicate increased ty. symbols suitability. aspect from elevation data were developed. In order to have the capability to estimate hab- maps prior to encoding. Some criterion maps itat conditions in the future, a separate computer were coded manually by overlaying a transparent program was developed to predict percentage veg- grid pattern and recording a value for each grid etative species composition and age distribution cell element. The value of an element in this of all tree species in stands as a result of topo- grid was the relative magnitude of the intensity graphic features, managerial actions, and natural of that criterion in that cell. Data were stored vegetative succession. This program used a Mar- within the program as a 120 by 120 matrix of in- kov chain (Phillips et al. 1976:232-242) ap- teger values, with each element (cell) represent- proach to forecast tree size-class distributions, ing an area of approximately 1.05 ha. and a Boolean approach to determine tree species Other criterion maps were coded semi- likely to regenerate successfully on a site based automatically using an electronic digitizer. A on the biotic conditions present. Although it is Numonics 237 Graphics Calculator/digitizer amenable to cellular data bases, the prediction (Numonics Corp., Lansdale, Pennsylvania) and a program is in need of further testing to become Tektronix 4051 minicomputer (Tektronix Corp., an operational part of the MAP4B system. Pend- Beaverton, Oregon) were used for this purpose. ing validation, results from the prediction COMPUTER-ASSISTEDHABITAT MAPPING * Williamson and WIelan 305

program may be used to predict future bear habitat suitability based on projected vegetative characteristics. ! RESULTS AND DISCUSSION A goal of this research was to provide informa- J tion pertinent for black bear management in T Shenandoah National Park. To this end, a series II of thematic maps was prepared for the McGaheysville Quadrangle, a 8857.6-ha test area in the southern portion of Shenandoah. Maps displaying habitat criteria used in the assessment procedure, and the location of Park facilities such as trails, campgrounds, and picnic areas, were produced along with a composite map of bear habitat suitability (see Fig. 1). The darker symbols, representing increasing habitat suitability, indicated that the most suitable habitat lies to- ward the Park interior in a somewhat dispersed

fashion...... Although contrasting symbols were nor- ...... *;.*,,,,, i mally assigned along equally-sized intervals using :::...... :: :: :: : :: :: i :::: ://///?::::, :.:: the greatest cumulative sum value within a map ...... as the upper limit, this feature was altered in or- I Z4S671/ 94.I2Z340678t12)4S& 412345S67' ,a .....;:::...... l2)145&71 )1 2)4S)4502 7|11 2 4567890 1 24567196 der to enhance visual contrast in Fig. 1. This 6.2m .1.033 . procedure was found particularly useful in isolat- S ing areas having superior or inferior as 4S0. 2. 01) potential 9 s. ..l3 bear habitat. *0.04,3 q, Areas of specific managerial interest were de- 3. 1.271 lineated also with 433 .ll the MAP4B 47. 2z system. To dem- O onstrate this application, a map was prepared of potential mating areas, i.e., sites with a high de- Fig. 2. Computer-generated map of potential bear breeding areas. Darker symbols indicate increased gree of remoteness from human disturbance and suitability. an abundance of cover. Examination of Fig. 2 will reveal an area having high potential suitabili- areas that ty as a mating habitat just left of the map center. lating may require special managerial Several other suitable areas were delineated near attention. The black bear habitat assessment procedure for Shenandoah the lower right corner of the map. It is notewor- was appropriate specifical- to Shenandoah National Park, and be nei- thy that the Skyline Drive courses through this ly may ther nor in latter area, and that virtually all of the better appropriate valid other localities or locations. mating habitat within this map was found to be However, the system is deliberately gen- eral and in within 2 km of the highway. Information of this may be used other localities, provided data are type could be useful in planning developments appropriate input available. The MAP4B was (e.g., hiking trails and campgrounds) that could system designed to assess habitat and not result in undue disturbance of bears by visitors. potential predict or estimate actu- al numbers of bears. The MAP4B system was not developed or designed to produce definitive CONCLUSIONS information; it was designed to convert habitat Although MAP4B has been implemented in data into reasonable estimates of habitat potential the Shenandoah Park only on an experimental based on predetermined and prespecified criteria. basis, the system appeared to be an effective In assessing bear habitat in Shenandoah Park, re- method for displaying spatial information and iso- alistic criteria, based on the best information 306 COMPUTER-ASSISTEDHABITAT MAPPING * Williamsonand W/elan

available, were used. However, much additional LITERATURECITED research on black bear ecology in Shenandoah BAPTISTE,M.E., J.B. WHELAN,AND R.B. FRARY. 1979. Visi- and elsewhere will be needed to validate totally tor perception of black bear problems at Shenandoah Na- the criteria used in assessing bear habitat in tional Park. Wildl. Soc. Bull. 7:25- 29. Shenandoah. BEEMAN,L.E. 1975. Population characteristics, movements, The primary utility of the MAP4B system is and activities of the black bear (Ursus americanus) in the the capabilityto allow the user to articulate and Great Smoky Mountains National Park. Ph.D. Thesis. Knoxville. define bear habitat potential based on known or Univ. Tennessee, 232pp. E.L. 1950. Deciduous forests of eastern North Amer- and to derive BRAUN, suspected ecological relationships, ica. Blakiston Co., Philadelphia, Pa. 596pp. quantitative and qualitative estimates of habitat Du BROCK,C.W. 1980. An analysis of Virginia black bear potential. However, the long-term utility of population dynamics. M.S. Thesis. Virginia Polytechnic MAP4B lies in its use as a heuristic habitat as- Inst. and State Univ., Blacksburg. 115pp. sessment model. This model may be improved FEDERATIONOF ROCKYMOUNTAIN STATES. 1977. Composite over time pending the addition of new informa- mapping system II users' guide. Federation of Rocky tion from experience or research. New habitat Mountain States, Inc., Denver, Colo. 209pp. J.W. 1976. A of black bear in criteria be added, or previous criteria be RAYBOURNE, study populations may may Virginia. Trans. Northeast Fish and Wildl. Conf. deleted or reduced in importancebased on chang- 33:71 -81. ing opinions of the user. Used in such a manner, PHILLIPS,D.T., A. RAVIDRAN,AND J.J. SOLBERG.1976. Op- MAP4B could be a powerful tool in the black erations research: principles and practice. John Wiley & bear management effort in ShenandoahNational Sons, Inc., New York, N.Y. 585pp. Park.