The timing of bird migration is fascinating to birders. We particularly look forward in the spring to our “first of the year” sightings of various migratory breeding birds.

The fall departure of those breeding birds is equally interesting but much more poorly documented.

The explanation for the difference is a simple one. When you see your first ovenbird of the spring, you record that date and you are done. Documenting the last departure requires continuous record-keeping. Perhaps you are able to go birding every day in late September. On Sept. 25 and 26, you see a gray catbird, none on Sept. 26 and 27, one on Sept. 28, and then none for the rest of the fall. You duly record Sept. 28 as the last fall departure date, but what an ordeal it was to get that date. Ideally, a birder seeking to document last departure dates should be out birding every day in the fall. As wonderful as that would be, who has the time?

To improve our understanding of the rhythm of Maine fall migration, I used the eBird database. This online application, administered by the Cornell Laboratory of Ornithology, is a boon for birders and researchers. The eBird application will keep track of your life lists and generate target lists for birds you haven’t seen yet.

Each birder’s sightings are available by request by ornithological researchers with approved research projects.

Between 1994 and 2017, I coordinated a volunteer-based spring arrival project, so I used the same years for the departure date project to complement it. EBird was not launched until 2002. Some birders have uploaded data prior to 2002 onto the site, but most of the 65,000 (!) records for this study came from 2002 and after.


I had adequate data to analyze 85 species of migratory breeding birds. To examine the tempo of migration, I divided the state into three equal latitudinal bands: the North Region, south to the latitudinal line through Jackman; the Central Region, south to the latitudinal line through Limerick; and the South Region, south to Kittery. We expect that migration will proceed from north to south, reflecting the harsher climate as one proceeds northward.

The eBird data I used are simply sightings of migratory birds, not necessarily last departure dates. To extract last departure dates, I had to define a beginning and ending point for each species.

I used the Last Safe Date from the Maine Breeding Bird Atlas to define the beginning of fall migration for each species in the study. Any observation later than the Last Safe Date cannot be used as evidence of breeding for a species, because fall migrants already will have started to pass through the state. The end of fall migration was defined as Dec. 31, although most species depart well before then.

I analyzed the latest 10 percent of dates for each species in each of the three regions. Of the 85 species analyzed, 64 conformed to the expected pattern: earliest departures from the North Region and latest departures from the South Region. Eleven species showed no difference between two regions.

Ten species showed surprising patterns. Spotted sandpipers departed last from the North Region. Nine species departed last from the Central Region. I think the explanation lies in the quality of the stopover habitat. Spotted sandpipers likely had better staging areas in Aroostook County than in points south. It was easy to miss spotted sandpipers in the south or central regions because they moved through so quickly.

Diet influenced migratory schedules. Aerial insectivores (swallows, swifts) were the first to leave, followed by leaf-gleaning insectivores (warblers, vireos), with seed-eaters (sparrows, buntings) bringing up the rear.  Seeds can be reliably found later in the fall after the insects have disappeared.

You can download a copy of my paper here.

Herb Wilson is an emeritus professor of biology at Colby College, where he taught ornithology and other courses. He welcomes reader comments and questions at [email protected]

Comments are not available on this story.