The prothonotary warbler is a long-distance migrant that occurs in the bottomland hardwood and floodplain forests of the Southeast. Because it nests in cavities and readily accepts nest boxes this species is well-studied.
Petit (1999) provides an excellent, detailed description of this species’ habitat requirements in his Birds of North America account:
“Key (and nearly universal) features are presence of water near wooded area with suitable cavity nest sites. Nest usually placed over or near large bodies of standing or slow-moving water, including seasonally flooded bottomland hardwood forest, baldcypress swamps, and large rivers or lakes (Walkinshaw 1953, Blem and Blem 1991). Many other forms of water also chosen, such as creeks, streams, backyard ponds, and even swimming pools. Water depth under nests highly variable. In Illinois, water depth below most nests 8–60 cm (n = 22; Kleen 1973). Nests located away from water are usually in low-lying, temporarily flooded spots (Walkinshaw 1953, LJP).
Other important habitat correlates include low elevation, flat terrain, shaded forest habitats with sparse understory, and in some places, presence of baldcypress (Kahl and others 1985, Robbins and others 1989). Common overstory trees in nesting habitat include willows, maples, sweet gum, willow oak, ashes, elms, river birch, black gum, tupelo, cypress, and other species associated with wetlands. Buttonbush is the most common subcanopy species. Canopy height 12–40 m (usually 16–20), canopy cover usually 50–75 percent; ground vegetation usually very sparse and of low stature (<0.5 m; Kahl and others 1985).
Exhibits area sensitivity, avoiding forests <100 ha in area and avoiding waterways with wooded borders <30 m wide (Kahl and others 1985).”
The model for prothonotary warbler habitat suitability contains seven parameters:
- successional age class
- forest patch size
- percent forest in the local (1-km radius) landscape
- snag density
The first suitability function combines landform, landcover, and successional age class into a single matrix (SI1) that defines unique combinations of these classes
. We directly assigned suitability index scores to these combinations based on the relative ranking of habitat associations reported by Hamel (1992) for prothonotary warblers in the Southeast.
Prothonotary warblers are rarely found >200 m from water during the breeding season. Therefore, we utilized a 9 × 9 window to examine whether water occurred close enough to each site to make it suitable (SI2). If water was present in any of the 81 pixels comprising the window, the center pixel was assigned a value of 1.000. If water was absent, the center pixel was assigned a 0.000
We also included forest patch size (SI3) as a variable in the habitat suitability model because prothonotary warbler abundance is lower in small isolated fragments and thin riparian buffer strips
. However, prothonotary warblers do occur in small forest fragments within heavily forested landscapes so we included percent forest in the local landscape as a model parameter (SI4). We fit a logistic function
to a dataset
reflecting the assumptions that landscapes with <30 percent forest were poor habitat (suitability index score ≤ 0.100) and landscapes with >70 percent forest were excellent habitat (suitability index score ≥ 0.900). We applied the maximum value of either SI3 or SI4 to all sites to increase the suitability of small forest blocks in predominantly forested landscapes above that which would be assigned based solely on patch size.
Prothonotary warblers are a cavity nester and utilize snags (SI5) as a nesting substrate. We assumed 5 snags/ha supplied an adequate number of cavities for nesting and roosting to make a site optimal habitat
. However, we recognize prothonotaries also utilize cavities in live trees as well as crevices for nest sites. Therefore, we assigned a residual suitability index score (0.250) to sites lacking snags. We fit a logistic function through these points to quantify the snag density-habitat suitability relationship
To calculate the overall SI, we calculated the geometric mean of the two suitability indices related to forest structure (SI1 and SI5) and the product of the maximum of the two SIs related to landscape composition (SI3, SI4) and SI2 separately and then the geometric mean of these values together.
Overall SI = ((SI1 * SI5)0.500 * (Max(SI3, SI4) * SI2))0.500