Modern views based on studies of bird behavior and physiology indicate migration is a regular, annually induced movement, modified by local weather conditions, but largely independent of them. Migration is a phenomenon far too regular to be created anew each season merely under stress of circumstances, such as need for food; and it begins before the necessity for a change in latitude becomes at all pressing. Swallows, nighthawks, shorebirds, and others may start their southward movement while the summer food supply in the North is at peak abundance. American robins and bluebirds may leave abundant food in the South and press northward when food supplies there are almost entirely lacking and severe cold and storms are likely to cause their wholesale destruction. Regularity of arrival and departure is one of the most impressive features of migration, and since birds travel in a rather strict accordance with the calendar, we might ask: "What phenomena, other than the regular changes in length of day, occur with sufficient precision to act as a stimulus for migration?"
Experimental work has abundantly demonstrated the effect of increased light upon the growth, flowering, and fruiting of plants. Similarly, Rowan's (1925) experiments with slate-colored juncos and the work of numerous subsequent investigators showed, at least in some temperate zone species of migratory birds, increasing periods of daylight triggered sex organs to develop, fat to be deposited, and migration restlessness to begin (King and Farner, 1963). When these conditions develop to a certain level, the bird enters a "disposition to migrate" and takes off for its breeding or wintering grounds. There is reason to believe certain weather conditions influence the actual time of departure and especially the rate of progress to the breeding area.
This explanation of the stimulus for migration may apply very broadly to birds that winter in temperate parts of the world and nest in the same hemisphere but fails in those birds wintering in the tropics, where little change in length of day occurs and even decreases during the spring in regions south of the Equator. It might be asked: "If the lengthening day is the stimulating factor, why should our summer birds, wintering in the tropics, ever start north?" In addition, if daylength influences when birds are stimulated to migrate, why should they not all leave the same locality at the same time? Or, if weather controls the departure of birds from a given area, should not all the migrants leave when conditions are optimal and refrain from departing when conditions are not so? Actually, the conditions that place a bird in a disposition to migrate are probably the result of a combination of factors affecting different species differently. Thus not all birds arrive at this condition at the same time.
It has been demonstrated experimentally that Andean sparrows, resident in equatorial regions, come into breeding condition twice annually entirely independent of changing light periods (Miller 1963); evidently the breeding cycle is controlled by periodic internal stimuli. Probably northern migrants that winter in equatorial regions and beyond have their migratory urges controlled by similar rhythms or biological clocks. Also, no evidence suggests that the southward migration of birds is controlled by changing periods of light even among species such as white-crowned sparrows, for which this is a controlling factor in the spring. The fall stimulus is probably an innate cyclic occurrence brought on by a biological mechanism of unknown nature (King, Barker, and Farner 1963).
It is pertinent to point out that the migratory instinct appears to be more or less transitory and not persistent over an extended period. Migratory birds may be delayed en route, either by natural conditions such as unusually abundant food supplies or forcibly by man. If detained until the end of the migratory season, migrants may not attempt to finish the journey because they apparently lose the migratory impulse. In the fall and early winter of 1929, abundant food and open water caused an unusual number of mallards to arrest their migration and remain in western Montana and northern Idaho. Later, however, when a heavy snowfall with subzero temperatures suddenly cut off the food supply, great numbers of the birds subsequently starved to death; a flight of a few hours could have carried them to a region of open water and abundant food.
WHEN BIRDS MIGRATE
One ordinarily thinks of the world of birds as sedentary during two periods each year, at nesting time, and in winter. For individuals this is obviously the case, but when the entire avifauna of North America or the world is considered, it is found that at almost all periods there are some latitudinal movements of birds. A few of these movements reoccur year after year with calendar-like regularity. Each species, or group of species, migrates at a particular time of the year and some at a particular time of the day. In this section some of the interesting differences will be discussed as to when birds migrate.
Time of Year
Some species begin their fall migrations early in July, and in other species distinct southward movements can be detected late into the winter. While some migrants are still traveling south, some early spring migrants can be observed returning north through the same locality. For example, many shorebirds start south in the early part of July, while the goshawks, snowy owls, redpolls, and Bohemian waxwings do not leave the North until forced to do so by the advent of severe winter weather or a lack of customary food. Thus an observer in the northern part of the United States may record an almost unbroken southward procession of birds from midsummer to winter and note some of the returning migrants as early as the middle of February. While on their way north, purple martins have been known to arrive in Florida late in January, and, among late migrants, the northern movement may continue well into June. In some species the migration is so prolonged that the first arrivals in the southern part of the breeding range will have performed their parental duties and may actually start south while others of the species are still on their way north.
A study of these facts indicates the existence of northern and southern populations of the same species that have quite different migration schedules. In fall, migratory populations that nest farthest south migrate first to the winter range because they finish nesting first. For example, the breeding range of the black-and-white warbler covers much of the eastern United States and southern Canada northwest through the prairies to Great Bear Lake in Canada ([Fig. 1.]). It spends the winter in southern Florida, the West Indies, southern and eastern Mexico, Central America, and northwestern South America. In the southern part of its breeding range, it nests in April, but those summering in New Brunswick do not reach their nesting grounds before the middle of May. (Lines that connect points where birds arrive at the same line are called isochronal lines. [Fig. 2.]) Therefore, if 50 days are required to cross the breeding range, and if 60 days are allowed for reproductive activities and molting, they would not be ready to start southward before the middle of July. Then with a return 50-day trip south, the earliest migrants from the northern areas would reach the Gulf Coast in September. Since adults and young have been observed on the northern coast of South America by August 21, it is very likely that they must have come from the southern part of the nesting area.