Small Whorled Pogonia Habitat Model
go to: USFWS Gulf of Maine Watershed Habitat Analysis
go to: Species Table
Feedback: We welcome your suggestions on improving this model!

Draft Date:
March 2001

Species:
Small whorled pogonia, Isotria medeoloides (a federally endangered terrestrial orchid).

Use of Study Area Resources:
The small whorled pogonia is rare but widely distributed, from southern Maine to Georgia and west as far as Michigan (Gleason and Cronquist 1991).

Habitat Requirements:
This herbaceous perennial propagates either from seed or from rootstock buds which overwinter and develop the next year's growth (vonOettingen 1992).   The USFWS Recovery Plan (vonOettingen 1992) gives the following description of habitat requirements: "...occurs on upland sites in mixed-deciduous or mixed-deciduous/coniferous forests that are generally in second- or third-growth successional stages. Characteristics common to most I. medeoloides sites include sparse to moderate ground cover in the species' microhabitat, a relatively open understory canopy, and proximity to features that create long-persisting breaks in the forest canopy. Soils at most sites are highly acidic and nutrient poor, with moderately high soil moisture values. Light availability could be a limiting factor for this species." Soil type, physiography, forest cover, and site disturbance history may be key distribution factors in Maine and New Hampshire (Sperduto 1993, Rawinski 1986 in Sperduto and Congalton 1996).

In the Northeast the small whorled pogonia typically occurs in acidic sandy loam having a fragipan (impervious layer), associated with the pH and moisture regime which the plant may require. Slope of the terrain where the orchids are found was typically 8 to 15%, or at the base of steep slope (Sperduto and Congalton 1996). Moisture regime is also affected by slope. The vegetative cover is usually mixed deciduous/coniferous, mid-successional forest with gaps in the canopy.

Model:
We modified the habitat model of Sperduto and Congalton (1996) to accommodate the environmental data sets available for our extensive study area. Our model had 5 elements: the geographic range over which small whorled pogonia occurs, suitability of soils, slope of the terrain, appropriate vegetative cover, and attribution of highest habitat value scores to locations of known occurrences. Occurrence information was provided by the State Natural Heritage programs of Maine, New Hampshire, and Massachusetts. The occurrence data were split into two sets; we used one subset of 43 points in the center of the range for model development and used the remainder to test the model.

The area in which habitat was modeled was restricted to the USDA Forest Service Ecological subunits (Keys et al. 1995) in which small whorled pogonia occurrences were mapped. These subunits very generally correspond to the county boundaries drawn in Sperduto and Congalton (1996).

Small whorled pogonia appears to be associated with the existence of an impervious soil layer within sandy loams. Such "pan" layers were identified from characteristics developed in discussions with NRCS soils scientists in Maine and Massachusetts, and from examination of the soils data coinciding with known occurrences in 2 counties in Maine and New Hampshire. NRCS expert opinion, generally verified by the data comparison, suggested selecting soils having layers with a moist bulk density => 1.5 gm/cc, and which were not a marine clay. Organic (peat, muck) soils were not included. The selected soils also had a moderate slope (=> 8), another aspect of the Sperduto and Congalton model. Digital soils data were available at a 1:24000 scale (SSURGO) for about 40% of the area of interest. For the remainder of the area we made the same selections from the 1:250,000 scale (STATSGO) data. For each of these larger polygons we summarized the component percentages and kept polygons only where > 49% of the components were of the preferred types.  Data from the two soils coverages were assigned unique scores. The SSURGO soils were scored 0.5 (suitable soils) or 0. The STATSGO scores were 0.3 (suitable soils) or 0. These data were combined by overlaying the higher on the lower resolution information. 

Slope and aspect were examined in relation to 43 small whorled pogonia occurrences.  Sperduto and Congalton (1996) found 90% of occurrences had slopes of 7.85% to 21.2% (between 5 and 12 degrees).  We developed slope data from USGS 1:250,000 DEM data and also found a highly significant association of our points with this range of slopes, compared to a set of random points in the range of species within our study area.  We also examined aspect, using directions largely corresponding to those having most occurrences in Sperduto and Congalton (1996).  Aspect did not display a significant association, again conforming with the findings of Sperduto and Congalton (1996). Slopes of 5 to 12 degrees were mapped as suitable; lesser and greater slopes were regarded as unsuitable (0).

We overlaid landcover information, and retained only the areas having deciduous or mixed forest as well as appropriate soils and slopes.

Finally,  to indicate our greater confidence in habitats mapped at localities with known whorled pogonia occurrences, we buffered these sites at a 1/4 mi radius and re-scored habitats derived from the above sequence 1.0, when within these buffers.

In summary, habitats were mapped within a specified range, and were scored as follows:

Suitability - 0-1

Environmental Characteristics
0.3 having suitable soil type based on 1:250000 data, suitable slope, and having deciduous or mixed forest cover
0.5 having suitable soil type based on 1:24000 data, suitable slope, and having deciduous or mixed forest cover
1.0 having any suitable soil type, suitable slope, deciduous or mixed forest cover, and containing or within 0.25 mile of known occurrence.

Model Testing: We compared the distribution of mapped habitat around a random set of 272 upland points in the small whorled pogonia range to habitat around the 38 occurrence reserved from the entire small whorled pogonia data set. Of the 38 sites, 32 had mapped habitat, while only 114 sites out of the 272 randomly distributed sites had habitat. The Chi-square was highly significant, indicating that the overall model predicts localities useful to this species.

Sources:
Gleason, H.A. and A. Cronquist. 1991. Manual of Vascular Plants of Northeastern United States and Adjacent Canada. New York Botanical Garden, Bronx, NY. 910 p.

Keys, J.E., Jr., J.C. Carpenter, S. Hooks, F. Koenig, W.H. McNab, W. Russell and W. Smith. 1995. Ecological units of the eastern United States - first approximation (map and booklet of map unit tables), USDA Forest Service. Atlanta, GA.

Sperduto, M.B. 1993. Use of a Geographic Information System (GIS) to Predict Potential Habitat for Isotria medeoloides (Pursh) Raf. in New Hampshire and Maine. M.S. thesis, University of New Hampshire, Durham, NH. 83 p.

Sperduto, M.B. and R.G. Congalton. 1996. Predicting rare orchid (small whorled pogonia) habitat using GIS. Photogramm. Eng. Remote Sensing 62(11):1269-1279.

Von Oettingen, S.L. 1992. Small whorled  Pogonia (Isotria medeoloides) Recovery Plan, First Revision. USFWS Region 5, Hadley, MA. 59 p.