The American woodcock breeds in early successional habitat throughout its range (Keppie and Whiting, Jr. 1994). Typically, these young forest stands occur on moist, uncompacted soils that allow woodcock to probe for earthworms, the birds’ preferred food (Steketee 2000). Equally important is an interspersion of the forest with openings that provide sites for both courtship displays and roosting (Sepik and Derleth 1993). Utilized openings in Maine were generally ≥1.2 ha (Dunford and Owen 1973). Given the affinity of woodcock for openings and early successional habitat, Sprankle and others (2000) recommended even-aged forest management in rotational blocks that ensure both habitat requirements are met.
Most available quantitative information on breeding habitat for woodcock comes from the Northeast, particularly Maine and Pennsylvania (Straw and others 1986, McAuley and others 1996). Shrub cover is generally high (75–87 percent; Morgenweck 1977), while overstory cover is typically moderate (50–64 percent; Dunford and Owen 1973, Gregg and others 2000). Nests occur in young forest stands (Morgenweck 1977). McAuley and others (1996) compared nest sites to random sites and found lower basal area and fewer coniferous saplings, but higher densities of deciduous saplings and shrub stems around nests sites. Young broods inhabit young to mid-age forest interspersed with openings; older broods occupy sites with greater basal area but fewer mature trees (Morgenweck 1977).
Many habitat factors have been associated with the presence of woodcock (Storm and others 1995, Klute and others 2002, Sprankle and others 2000). Landcover variables were the best predictors at fine scales, whereas indices of landscape heterogeneity were the most important predictors at large spatial scales. Murphy and Thompson (1993) developed a model to predict density of males on singing grounds in central Missouri that contained small (≤2.5 cm d.b.h.) stem density, tree (>2.5 cm d.b.h.) density, and field size as predictor variables.
The American woodcock HSI model contains seven factors:
- SI1: Landform, landcover, and successional age class
- SI2: Small (<2.5 cm d.b.h.) stem density
- SI3: Composition of appropriately-sized foraging-nesting and courtship-roosting habitat patches in the landscape
- SI4: Soil moisture
- SI5: Soil texture
The first suitability function combines landform, landcover type, and successional age class into a single matrix (SI1) defining unique combinations of these classes
. Because woodcock prefer moist habitats with high deciduous stem densities, we assigned the highest suitability index scores to shrub-seedling-aged transitional, deciduous, and woody wetland cover types in floodplain-valley landforms. We considered mixed and evergreen forests, as well as xeric-ridge landforms poor quality American woodcock habitat.
We included small stem density (SI2) as a model function because woodcock rely on vertical structure to provide security from predators as they forage, nest, and loaf during the day. Estimates of stem density at nest sites range from 3767 stems/ha in Missouri to 49250 stems/ha in Pennsylvania (reviewed in McAuley and others 1996). We assumed the stem density value for Missouri represented usable but not ideal habitat for woodcock and associated it with a suitability index score of 0.250. Further, we assumed the stem density value for Pennsylvania was characteristic of ideal nesting habitat (i.e., suitability index score = 1.000) and higher stem densities were not detrimental to woodcock
. We fit a logistic function through these data points to quantify the small stem density – habitat suitability index score relationship
The next two factors in the model relate to the minimum size of habitat patches used by American woodcock. Movement rates within diurnal foraging and nesting habitats are often low, resulting in small diurnal home ranges (≤0.3 ha; Hudgins and others 1985). Conversely, woodcock display and roost in relatively large openings at night (≥1.6 ha; Keppie and Whiting, Jr. 1994, McAuley, pers. comm.). We used these data to establish minimum area thresholds for forests and openings, respectively. The ultimate suitability of either of these habitat types, though, is related to their interspersion with one another, as both are required by woodcock. Ideally, these habitats should be separated by <400 m (Hudgins and others 1985), even though average home range size may ≥74 ha (485 m radius; Keppie and Whiting, Jr. 1994). Because home ranges may encompass areas of non-habitat, woodcock can sometimes occur where the proportion of these habitat types within a typical home range is relatively small (e.g., 0.1; table 13). We assumed woodcock derive greater benefit from increasing proportions of early successional forest habitat than field habitat within their home ranges due to greater foraging opportunities and increased protection from predators. Thus, our table defining the relationship between landscape composition (SI3) and suitability index scores exhibits greater increases in suitability with relatively modest increases in diurnal habitat compared to the increases in suitability associated with similar proportional increased in openings.
Soil properties also influence American woodcock habitat suitability. Woodcock feed almost exclusively on earthworms, which they probe for preferentially in moist loamy soils (Rabe and others 1983). Soils with too much clay or sand do not contain enough accessible earthworms to support a foraging woodcock. Therefore, we included both soil texture (SI4) and soil drainage class (SI5) as variables in the woodcock habitat suitability model. We used the STATSGO database to define soil characteristics. Soil texture classes from STATSGO were cross-walked to soil texture classes from the soil triangle
and then assigned suitability index scores based on texture descriptions in Rabe and others (1983; table 15). Additionally, we assumed soil drainage class was associated with soil moisture content and similarly assigned suitability index scores to these drainage classes based on data from Rabe and others
To calculate the overall suitability index score, we determined the geometric mean of SI scores for forest structure (SI1 and SI2) and landscape factors (SI3, SI4 and SI5) separately and then the geometric mean of these means together.
Overall SI = ((SI1 * SI2)0.500 * (SI3* SI4 * SI5)0.333)