The Rich Hole Wilderness

A Computer Map of Potential Sites for the Small Whorled Pogonia (Isotria medeoloides) in Prince William Forest Park (A National Park)

Gary Casabona and Robert H. Giles, Jr.

Department of Fisheries and Wildlife Sciences Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061-0321

Gary Casabona gcasabona@NJ.USDA.GOV

Abstract: A computer map was made of the potentially suitable areas where the small whorled pogonia isotria medeoloides might be found. The study area was the Prince William Forest Park, Virginia, of the U.S. National Park Service. The mapping allows knowledge to be integrated and the map product may assist in searches, protection from disturbance, and recovery efforts.

Key Words: rare, threatened, endangered, GIS, computer maps.

The small whorled pogonia (Isotria medeoloides) is a rare orchid (Williams and Williams 1983) occurring in Virginia (Van Alstine et al. 1996). Its distribution and ecology is under study (U.S. Fish and Wildlife Service 1992; Vitt 1991; Ware 1991; Merhoff 1983, 1988, 1989; Keeman 1988). We sought procedures for aiding in synthesizing present knowledge about the plant, further understanding its ecology, and for improving sampling strategies. Knowledge of the ecology is essential to efforts to maintain species richness in National Park Service areas such as Prince William Forest Park of northern Virginia (Anon. 1991) and other areas throughout the state (Ware 1991).

We employed the GRASS geographic information system (GIS) (US CERL 1993) to map areas probably suitable for the occurrence of the plant (Casabona 1994). Because knowledge about the plant is limited (as it typically is for other threatened or endangered species) we sought maximum map discrimination with the fewest factors included in map layers. Previously a GIS analysis had been conducted to estimate potential habitat for the plant in New Hampshire and Maine (Sperduto 1993). The small whorled pogonia is extremely rare, existing only in a few scattered colonies in Michigan, New York, Pennsylvania, Vermont, New Hampshire, Maine, Massachusetts, Connecticut, New Jersey, Virginia, North Carolina, and Missouri. Since the plant may lie dormant for several years without flowering (Williams and Williams 1983), locating additional extant colonies is likely to be difficult. A Recovery Plan for the species has been drafted (US Fish and Wildlife Service 1992).

We hypothesized that through site analyses of soil type, insulation, slope, aspect, erosion, rainfall, and presence of other flora, appropriate areas for action such as reintroducing, preserving, studying, and managing particular taxa can be identified.

Map coverages of slope and aspect were generated from digital data on elevations such as that available as a Digital Elevation Model (DEM), (U.S. Geological Survey, Washington, DC), using various existing methods for their estimation. The GRASS program module called r.slope.aspect was used to produce slope and aspect maps, also called layers, or coverages. A raster map coverage of soil type was obtained from the Prince William Forest Park staff database.

Several colonies of the small whorled pogonia are known to exist within the Park. The exact locations of these colonies are being kept confidential by Park personnel to protect the plant and prevent disturbance.

The plant is found mostly in upland sites with mixed deciduous or mixed deciduous and conifer forest in second- or third-growth successional stages (Mehroff 1988). Soils are generally highly acidic and nutrient-poor with moderately high moisture. Sparse to moderate ground cover in its microhabitat allows light to penetrate to the short (10-25 cm) plants as does proximity to features that create breaks in the canopy (USFWS 1992).

GIS analyses for other threatened or endangered flora may need to address the relationship between plant and pollinator. Since the plant is self-pollinating (Mehroff 1983), such relationships were not germane to this analysis.

To create a map showing areas suitable for the plant, we reclassified the soil coverage according to the Soil Survey of Prince William Forest Park (Virginia Polytechnic Institute and State University 1979). We excluded areas having soils with a pH greater than 5.1 in the ('A' horizon of the soil based on the work of Mehroff (1988). Mehroff studied 9 populations of the plant and found that the soil pH was no higher than 5.1 at any of the sites. Since most populations are found in mesic to dry-mesic deciduous woodlands (Mehroff 1988), we also excluded very wet areas even if they had the appropriate soil pH. This soil classification step resulted in excluding 3025 ha (7,474 acres) representing 48.9% of the total area of the Park.

Since the plants are known to be associated with deciduous or mixed deciduous and conifer types, we excluded all areas on the landcover layer that did not fit these criteria. Since the plants are most closely associated with deciduous woodlands in the presence of deep leaf litter (Williams and Williams 1983), we excluded all area of mixed deciduous and conifer trees. These are typically transitional areas, progressively losing conifers. We created a cross product of the soil and landcover maps (GRASS program "r. cross") which identified the areas that had not been excluded in either coverage. Only 1,762 ha (4,354 acres) representing 28% of the area inside the Park was potentially suitable habitat for the small whorled pogonia based upon soil acidity and forest cover type alone. See Table 1.

To discriminate further within the remaining suitable areas. We attempted to develop a canopy density layer indicative of light penetration (Mehroff 1988; USFWS 1992). Since natural canopy breaks often occur near streams, we used

Table 1. Potentially suitable habitat for the small whorled pogonia based on soil acidity and forest covertype.

Soil Acidity	Covertype	Hectares	Acres	Percent Cover
Suitable	Suitable	1762	4354	28.48
Suitable	Unsuitable	1401	3462	22.64
Unsuitable	Suitable	1884	4655	30.45
Unsuitable	Unsuitable	1141	2819	18.43

the vector coverage of streams in the Park and mapped a "zone of influence" of 30 m x 30 m on each side of all streams.

Figure 1. Habitat suitability for Isotria medeoloides.

Using the map calculator program (r.mapcalc) stream (water) cells were subtracted from the map showing streams plus their zone of influence to give a map coverage of only 30- m-wide cells adjacent to streams. By crossing this coverage with the map showing potentially suitable habitat (Fig.1), we identified areas which satisfied the 3 criteria: proper soil acidity, related forest cover type, and proximity to canopy breaks (Fig.2).

Fig.2. The areas meeting the 3 criteria, shown darkest (black) on the map represent a total of about 26.2 ha (64.8 acres), only about 0.4% of the area within the Park.

Using the GRASS program "r. clump", we found that there were 156 patches of potential habitat for the plant. These patches have an average size of 1,684 m2. The chances of locating previously unknown colonies of the small whorled pogonia may be significantly higher if Park personnel concentrate their searches only in these cells. Mehroff (1989) found no colonies in areas having slopes greater than 30%. Of the 156 patches of highly suitable habitat in the Park, 1 patch, having a size of 1 cell had such a steep slope. No further exclusions, based on slope, were used in our study of the area.

GIS programs that calculate slope and aspect from digital elevation models can produce inaccuracies, especially when the study area is relatively flat, as is the case with Prince William Forest Park. It is unlikely that such errors would significantly affect the analysis. However, research personnel attempting to locate new colonies of Isotria medeoloides in the areas identified by the map should be aware that small patches on the landscape may be too steep to provide suitable habitat for the small whorled pogonia. Some error is inherent in GIS work. The relatively small cell size (30 m) of the data used in this analysis implies that the results should be accurate enough for many purposes of rare species study and management. However, this also assumes that the soil and landcover coverages are accurate enough to support the analysis at this resolution.

We contacted staff of the Prince William Forestry Park and presented the maps. Staff were appropriately protective of the plant. They reported that they were unaware of any plant colonies being in the cells selected as of high probability of occurrence. They did report occurrence in the second-most likely cells. A staff member of the Park confirmed that the map did show at least 2 areas where she had seen the plant. Invalidation (Holling 1978) remains the challenge, for intensive surveys have not been made and, as noted, the plants flower only for short periods, not every year, and are difficult to see.

Acknowledgement

The research presented in this article was accomplished in fulfillment of contract PX3700-2-0142 between Virginia Polytechnic Institute and State University and USDI, National Park Service, Prince William Forest Park, Triangle, VA 22172.

We appreciate the advice of Donna M. E. Ware, Dr. Richard Oderwald, and Dr. Paul L. Angermeier, Dr. Duncan M. Porter and especially Ms. Marie Frias of the National Park Service, Prince William Forest Park. A review of the manuscript by Carol A. Polia of the Park was helpful.

Literature Cited

Anonymous. 1991. General management plan (Prince William Forest Park). U.S. Department of the Interior, National Park Service. Triangle, Va. 14 pp.

Casabona, G. 1994. GIS procedures for analyzing wildlife topics in a National Park in Virginia. M.S. Thesis, VPI and SU, Blacksburg, VA. 99 pp.

Holling, C. S. (Ed.) 1978. Adaptive environmental assessment and management. International Inst. for Applied Systems Analysis. John Wiley and Sons, New York. xvii + 377 pp.

Keeman, P. E. 1988. Progress report on Isotria medeoloides. Am. Orchid. Soc. Bull. 57:624-626.

Mehroff, L. A. 1983. Pollination in the genus Isotria. Am. J. Bot. 70:1444-1453.

Mehroff, L. A. 1988. Reproductive vigor and environmental factors in populations of an endangered North American orchid, Isotria medeoloides (Pursh) Rafinesque. Biol. Conserv. 47:281-296.

Mehroff, L. A. 1989. The dynamics of declining populations of an endangered orchid, Isotria medeoloides. Ecology , 70:783-786.

Sperduto, M. B. 1993. Use of a GIS to predict potential habitat for Isotria medeoloides (Pursh) Raf. in New Hampshire and Maine. M.S. Thesis, University of New Hampshire, Durham, New Hampshire. 106 pp.

USA CERL. 1993. GRASS, Version 4.1 user's reference manual. Open Grass Foundation. Boston, Mass. 563 pp.

U.S. Fish and Wildlife Service. 1992. Small whorled pogonia (Isotria medeoloides) recovery plan, first revision. Newton Corner, MA. 75 pp.

Van Alstine, N. E., W. H. Moorhead, III, Allen Belden, Jr., T. J. Rawinski, and J. C. Ludwig. 1996. Recently discovered populations of small whorled pogonia (Isotria medeoloides) in Virginia. Banisteria 7:3-7.

Vitt, P. 1991. Conservation of Isotria medeoloides: a federally endangered terrestrial orchid. M.S. Thesis, University of Maine, Orono, Maine. 40 pp.

Ware, D.M. 1991. Small whorled pogonia, p. 95-97 in K. Terwilliger (coordinator) Virginia's Endangered species . Nongame and Endangered Species Program, Virginia Dept. of Game and Inland Fisheries. McDonald and Woodward Pub. Co., Blacksburg, Va.

Williams, J. G. and A. E. Williams. 1983. Field guide to the orchids of North America. Universe Press. New York, N.Y. 237 pp.

This Web site is maintained by R. H. Giles, Jr.
Last revision April 30, 2001.