Wood Thrush breeds throughout eastern forests, occurring east of the Great Plains from central Florida and east Texas north to southern Canada. The map lights up with their spring arrival in April and May.
Focusing on the latter half of the year for Wood Thrush raises an interesting challenge: when in fall does Wood Thrush migrate? After spring arrival, high density Wood Thrush breeding areas are bright, showing comparatively high frequency of occurrence. Over the course of the next two months, the bright areas of the map gradually become duller, which is an indication of declining detectability. Detectability follows regular seasonal patterns (see Northern Cardinal), and in Wood Thrush detectability is at its peak in late May and early June when birds are singing, but declines as these birds build nests and focus more energy on raising young. The detectability declines further in August, when most Wood Thrushes are molting, and it is at this time of year when the species is hardest to find in your local woodlands.
A century of bird study in eastern North America, including banding studies, reveals that Wood Thrush does not initiate migration until later. In fact, the actual southward movement of Wood Thrush, which peaks from early September to mid-October, is very hard to detect on this annual occurrence animation (there is a brief period when the map gets bright in the central Appalachians). Thrushes are particularly secretive forest birds, and when they are not vocalizing, they are very hard to find. So a comparison of the Wood Thrush map to those for more conspicuous migrants like Western Tanager highlights these differences in detectability as they relate to the different behaviors of the birds. This comparison also highlights the challenges in using these models to understand where and when migration is occurring, or where changes in occurrence may be more related to behavioral changes. Thrushes are so secretive on migration, that any signal south of breeding areas (such as on the Gulf Coast or in south Texas) is lost in comparison to the comparatively high occurrence in summer. Again, this does not mean that the model does not ‘understand’ their occurrence here, just that the scaling is such that it does not approach the levels of occurrence in summer. In future versions of these models we may have more flexibility to adjust the scale and ‘see’ the migration, even if the overall occurrence level is far below that for summer.