Chapter III: Part 3

The white-winged scoter, which also breeds in the interior of the continent from northern North Dakota north to the Arctic coast, was at one time credited with an elliptical migration route, at least insofar as those wintering on the Atlantic coast are concerned. This sea duck nests only near fresh water but spends the winters on the ocean along both the Atlantic and Pacific coasts of the United States. It migrates over land surfaces mainly at night and it is now believed that after working northward to the waters of Long Island Sound it starts up the valleys of the Hudson and Connecticut Rivers and flies overland to the Great Lakes, from which region it continues west and northwest to the breeding grounds, returning to its winter quarters over the same route. Early ideas alleging an elliptical route probably arose from the fact that great numbers of first-year nonbreeding birds regularly pass up the New England coast, cross the Gulf of St. Lawrence and spend the summer loafing off the coast of Labrador. In the fall these birds form into large flocks and retrace their flight, chiefly during daylight hours, to winter quarters from southeastern Maine south at least to Chesapeake Bay. As it was not known that the white-winged scoter does not usually breed until it is 2 years old, and since the southward movement of yearling birds was conspicuous while the travels of those from the nesting grounds were chiefly at night, the theory was advanced that the latter flew 1,500 miles due east from the region west of Hudson Bay to the coast of Labrador, thence southward to the known winter quarters.

A study of the Canada geese that winter abundantly in the waters of Back Bay, Va., and Currituck Sound, N. C., reveals another important tributary to the Atlantic coast route. Banding has shown that the principal breeding grounds of these birds are among the islands and on the eastern shore of Hudson Bay. From this region they move south in autumn to the point of Lower Ontario between Lakes Erie and Huron. Occasionally one of these geese will be recovered in the Mississippi Valley but the great majority are retaken either on their breeding grounds or on the Atlantic coast south of Delaware Bay, showing another instance of a long cross-country flight by waterfowl. Although the Canada goose is abundant in migration on the coast of New England, the birds taken there rarely include any of those banded in southern Ontario. The northeastern population of these geese comes from breeding areas in New England, the Maritime Provinces of Canada, Newfoundland, and the desolate coast of Labrador, their migration being entirely coastwise. Still another cross-country route between the Mississippi Valley and the Atlantic coast remains to be briefly described. It is not yet well understood, but the banding of ducks such as the blue-winged teal on the coastal saw-grass marshes of South Carolina has revealed that there is a migration route across the Appalachians to the Mississippi Valley. Birds marked in these marshes have been retaken in Tennessee and Kentucky, as well as in other States farther north in the Mississippi Flyway.

=Mackenzie Valley--Great Lakes--Mississippi Valley route and tributaries=

Easily the longest route of any in the Western Hemisphere is that extending from the Mackenzie Valley past the Great Lakes and down the Mississippi River, including its tributaries. Its northern terminus is on the Arctic coast in the regions of Kotzebue Sound, Alaska, and the mouth of the Mackenzie River, while its southern end lies in Patagonia. During the spring migration some of the shore birds traverse the full extent of this great path, and it seems likely that the nighthawk, the barn swallow, the blackpolled warbler, and individuals of several other species that breed north to Yukon and Alaska must twice each year cover the larger part of it.

For more than 3,000 miles--from the mouth of the Mackenzie to the delta of the Mississippi--this route is uninterrupted by mountains. In fact, there is not even a ridge of hills high enough to interfere with the movements of the feathered travelers, and the greatest elevation above sea level is less than 2,000 feet. Well timbered and watered, the entire region affords ideal conditions for the support of its great hosts of migrating birds. It is followed by such vast numbers of ducks, geese, shore birds, blackbirds, sparrows, warblers, and thrushes, that observers stationed at favorable points in the Mississippi Valley daring the height of migration can see a greater number of species and individuals than can be noted anywhere else in the world.

Starting in the region of Kotzebue Sound, Alaska, the route extends eastward across northern Alaska and joins another that has its origin at the mouth of the Mackenzie River. The line of flight then trends a little east of south through the great lake system of central Canada, where it is joined by two or three other routes from the northeast that have their origin on the central Arctic coast. Continuing southward the migrating flocks are constantly augmented by additions to their numbers as they pass over the great breeding grounds of central and southern Canada. Upon reaching the headwaters of the Missouri and Mississippi Rivers the route follows these streams to the Gulf coast, Arriving in this latitude many species, including the ducks and geese, the robin, the myrtle warbler, and some others spread out east and west for their winter sojourn. Others, despite the perils of a trip involving a flight of several hundred miles across the Gulf of Mexico, strike out boldly for Central America and South America. This part of the route is a broad "boulevard" extending from northwestern Florida to eastern Texas and reaching southward across the Gulf of Mexico to Yucatan and the Isthmus of Tehuantepec (fig. 17, route 4). In other words, when most of the birds (chiefly shore birds and land birds) that have come south directly through the Mississippi Valley have reached the coastal plains of Mississippi and Louisiana, they continue directly across the Gulf of Mexico. This route appears to have preference over the safer but more circuitous land or island routes by way of Texas and Florida. During the height of migration some of the islands off the coast of Louisiana are wonderful observation points for the student of birds, as the feathered travelers literally swarm over them.

Present detailed knowledge of the chief tributaries to the Mackenzie--Great Lakes--Mississippi Valley highway relates chiefly to waterfowl. Reference already has been made to the flight of the black ducks (p. 58) that reach the Mississippi Valley from southern Ontario. Some individuals of this species banded at Lake Scugog, Ontario, have been recaptured in succeeding seasons in Wisconsin and Manitoba, but the majority have been retaken at points south of the junction of the Ohio River with the Mississippi, definitely indicating their route of travel from southern Ontario.

A second route that joins the main artery on its eastern side is the one used by the blue goose, the migration route of which is probably more nearly due north and south than that of any other North American bird. The breeding grounds, which only in recent years have been discovered, are mainly in the Fox Basin region of Baffin Island and on Southampton Island. In fall these geese work southward, chiefly along the eastern shore of Hudson Bay, and upon reaching the southern extremity of James Bay they take off for what is practically a nonstop flight to the great coastal marshes of Louisiana west of the delta of the Mississippi River. In some seasons the flocks make intermediate stops among the islands and sand bars of the Mississippi, as they are occasionally common in the general vicinity of Memphis, Tenn. Most of the birds push on, however, and during the period from the first of November to the last of March fully 90 percent of the species are concentrated in the area between the Sabine and the Mississippi Rivers. On the return trip northward there is sometimes a tendency for some of the blue geese to veer off toward the northwest, as they are occasionally abundant in eastern South Dakota and southeastern Manitoba. It is of particular interest to note that while some other geese and many ducks start their northward journey at the first sign of awakening spring, the blue goose remains in its winter quarters until the season there is far advanced, seemingly aware that its own breeding grounds in the Arctic are still in the grip of winter.

=Great Plains--Rocky Mountain routes=

A great western highway also has its origin in the Mackenzie River delta area and in Alaska. This is used chiefly by the pintail and the American widgeon or baldpate, which fly southward through eastern Alberta to western Montana. Some localities in this area, as for example, the National Bison Range at Moiese, Mont., normally furnish food in such abundance as to induce these birds to pause in their migratory movement. Upon resuming travel, some flocks move almost directly west across Idaho to the valley of the Columbia River, from which they turn abruptly south to the interior valleys of California. Others leave the Montana feeding and resting areas and turn southeastward across Wyoming and Colorado to join the flocks that are moving southward through the Great Plains (fig. 15).

Many redheads that breed in the Bear River marshes in Utah take a westerly route across Nevada to California, but some leave these breeding grounds and fly northeastward across North Dakota and Minnesota to join the flocks of these ducks that come out of the prairie regions of Canada. A few of them even travel southeastward to the Atlantic coast. This route can be traced by the records of ducks banded in summer in the Bear River marshes and retaken the following fall at points in eastern Montana, Wyoming, South Dakota, North Dakota, Minnesota, Wisconsin, Michigan, Ohio, and Maryland. Great numbers, however, follow another route from these marshes across the mountains in an easterly direction, where it almost immediately turns southward through Colorado and New Mexico, and continues to winter quarters in the Laguna Madre off the coast of Texas or in the Valley of Mexico (fig. 17, route 6). This route also represents the travels of many of the land birds of the Rocky Mountain region. Such birds perform comparatively short migrations, most of them being content to stop when they reach the middle districts of Mexico, only a few passing beyond the southern part of that country.

Observations made in the vicinity of Corpus Christi, Tex., have shown one of the short cuts (fig. 17, route 5) that is, in effect, a part of the great artery of migration. Thousands of birds pass along the coast to the northern part of the State of Vera Cruz. As the coast of the State of Tamaulipas to the north is arid and so entirely unsuited to the needs of birds that are frequenters of moist woodlands, it is probable that much or all of this part of the route is a short distance off shore. It is used by such woodland species as the golden-winged warbler, the worm-eating warbler, and the Kentucky warbler.

=Pacific coast route=

Although it does present features of unusual interest, the Pacific coast route is not of so great importance as some of the others described. Because of the equable conditions that prevail, many species of birds along the coast from the Northwestern States to southeastern Alaska either do not migrate at all or else make relatively short journeys. This route has its origin chiefly in Alaska, the general region of the delta of the Yukon River marking its northern terminus, although a few species join it after a flight westward along the Alaskan Arctic coast. Some of the scoters and other sea ducks of the north Pacific region, and the diminutive cackling goose which breeds in the delta of the Yukon River, use the coastal sea route for all or most of their southward flight. The journey of the cackling goose, as shown by return records from birds banded at Hooper Bay, Alaska, has been traced southward across the Alaska Peninsula and apparently across the Gulf of Alaska to the Queen Charlotte Islands, the birds following the coast line south to near the mouth of the Columbia River. There the route swings toward the interior for a short distance before continuing south by way of the Willamette River Valley. The winter quarters of the cackling goose are chiefly in the vicinity of Tule Lake, on the Oregon-California line, and in the Sacramento Valley of California, though a few push on to the San Joaquin Valley.

A tributary of this flyway is followed by Ross's goose which is now known to breed only in the Perry River district south of the Queen Maud Gulf on the central Arctic coast of Canada (fig. 20). Its fall migration appears to be southwest and south across the barren grounds to Great Slave and Athabaska Lakes, where it joins thousands of other waterfowl bound for their winter homes along the eastern coast of the United States and the Gulf of Mexico. But when the Ross's geese have traveled south approximately to the northern boundary of Montana, they separate from their companions, and turning to the southwest cross the Rocky Mountains and settle for the winter in California.

The route taken by the white-winged scoters that winter on the Atlantic coast already has been indicated (p. 59). Some birds of this species, however, winter on the Pacific coast from Puget Sound south to southern California. Their passage by thousands up and down the coast has been noted as far north as northwestern British Columbia. The species is known to nest in Alaska, which may be the home of some, at least, of the scoters that winter on the Pacific coast. If such be the case, however, it must be admitted that a part of the route taken by the birds when on migration is unknown, though very few observations are available from the interior of northern British Columbia, across which the route may lie.

The southward route of those migratory land birds of the Pacific coast that in winter leave the United States extends chiefly through the interior of California to the mouth of the Colorado River and on to winter quarters in western Mexico.

The movements of the western tanager show a migration route that is in some ways remarkable. The species breeds in the mountains from the northern part of Baja California and western Texas north to northeastern British Columbia and southwestern Mackenzie. Its winter range is in two discontinuous areas--southern Baja California and eastern Mexico south to Guatemala (fig. 21). On the spring migration the birds enter the United States about April 20, appearing first in western Texas and the southern parts of New Mexico and Arizona (fig. 22). By April 30 the vanguard has advanced evenly to an approximate east-and-west line across central New Mexico, Arizona, and southern California. But by May 10 the easternmost birds have advanced only to southern Colorado, while those in the far west have reached northern Washington. Ten days later the northward advance of the species is shown as a great curve, extending northeastward from Vancouver Island to central Albert and thence southeastward to northern Colorado. Since these tanagers do not reach northern Colorado until May 20, it is evident that those present in Alberta on that date, instead of traveling northward through the Rocky Mountains, which from the location of their summer and winter homes would seem to be the natural route, reached there by the Pacific coast route to southern British Columbia and thence across the mountains, despite the fact that these are still partly covered with snow at that time.

=Pacific oceanic route=

The route of the Pacific golden plover is fully as interesting and as remarkable as the elliptical course followed by its eastern cousin (fig. 18). The breeding range of the eastern golden plover extends through Arctic America west to the northern coast of Alaska where, in the vicinity of Point Barrow, it meets the nesting grounds of the Pacific form, which is really an Asiatic subspecies. It breeds chiefly in the Arctic coast region of Siberia and merely overflows onto the Alaskan coast, some of the birds probably migrating south along the coast of Asia to winter quarters in Japan, China, India, Australia, New Zealand, and Oceania, including the Hawaiian Islands, the Marquesas Islands, and the Low Archipelago. Golden plovers in migration have been observed at sea on a line that apparently extends from these islands to the Aleutians, and it therefore appears certain that at least some of the Alaskan birds make a nonstop flight across a landless sea from Alaska to Hawaii. While it would seem incredible that any birds could lay a course so straight as to attain these small oceanic islands, 2,000 miles south of the Aleutians, 2,000 miles west of Baja California, and nearly 4,000 miles east of Japan, the evidence admits only the conclusion that year after year this transoceanic round-trip journey between Alaska and Hawaii is made by considerable numbers of golden plovers.

The Pacific oceanic route probably is used also by the arctic terns that breed in Alaska, and possibly by those from the more western tern colonies of Canada. This species is of regular occurrence on the western coasts of both the United States and South America, indicating that the western representatives travel southward to the Antarctic winter quarters without the spectacular migration features that appear to characterize the flight of those from the eastern part of the continent (fig. 8).

=Arctic routes=

In the discussion of the migration of the Arctic tern (p. 38) it was noted that this species makes a very distinct west-to-east movement across northern Canada, continuing the flight eastward across the Atlantic Ocean toward the western coast of Europe. It seems likely that there are other species, including the parasitic jaeger, that regularly breed in the northern part of the Western Hemisphere but migrate back to the Old World for their winter sojourn. Some others, as the red-legged kittiwake and Ross's gull, remain near the Arctic region throughout the year, retreating southward in winter only a few hundred miles. The emperor goose in winter is found only a relatively short distance south of its breeding grounds, and eider ducks, although wintering in latitudes well south of their nesting areas, nevertheless remain farther north than do the majority of other species of ducks.

The routes followed by these birds are chiefly coastwise, and in the final analysis may be considered as being tributary either to the Atlantic or to the Pacific coast routes. The passage of gulls, ducks, the black brant, and other water birds at Point Barrow, Alaska, and at other points on the Arctic coast, has been noted by several observers, and from present knowledge it may be said that the best defined Arctic route in North America is the one that follows around the coast of Alaska.

_Evolution of Migration Routes_

From the foregoing descriptions of migration routes it will be observed that the general trend of migration in most species of North American birds is northwest and southeast. It is comparatively easy to trace the probable steps in the evolution of the migrations of some species, and some routes have developed so recently that they still plainly show their origin.

The tendency is for eastern species to extend their ranges by pushing westward, particularly in the north. For example, in the Stikine River Valley of northern British Columbia and southwestern Alaska the eastern nighthawk, eastern chipping sparrow, rusty blackbird, eastern yellow warbler, redstart, and others have established breeding stations at points 20 to 100 miles from the Pacific Ocean. The robin, flicker, slate-colored junco, blackpolled and myrtle warblers, and ovenbird, all common eastern species, also are established as breeding birds in western Alaska, the ovenbird having been detected on the lower Yukon River. These birds are essentially Atlantic and Mississippi Flyway species, however, and so do not migrate in fall by any of the Pacific or Central routes, but instead retrace their journey across the mountains and move southward along the broad flyways of the East.

The red-eyed vireo, a striking example of an abundant woodland bird, is essentially an inhabitant of States east of the Great Plains, but an arm of its breeding range extends northwest to the Pacific coast in British Columbia (fig. 23). It seems evident that this is a range extension that has taken place comparatively recently by a westward movement from the upper Missouri Valley, and that the invaders retrace in spring and fall the general route by which they originally entered the country.

In the case of the bobolink, a new extension of the breeding range, and a consequent change in the migration of the species, has taken place since the settlement of the country (fig. 19). A bird of damp meadows, it was originally cut off from the Western States by the intervening arid regions. But with the advent of irrigation and the bringing of large areas under cultivation, small colonies of nesting bobolinks have appeared at various western points, and now the species is established as a regular breeder in the great mountain parks and irrigated valleys of Colorado and elsewhere almost to the Pacific coast. In retracing their course to reach the western edge of the route followed by the bulk of the bobolinks that breed in the northern United States and southern Canada, these western pioneers must fly long distances along a line that runs almost due east and west.

Similarly it is possible to sketch what seems to be the logical evolution of the remarkable routes of the golden plover (fig. 18). It may be assumed that the eastern birds of this species first followed an all-land route from the South American winter quarters through Central America, Mexico, and Texas to the western parts of the Mississippi Valley. As the migration route lengthened northward with the retreat of the ice and the bird's powers of flight developed, there would be a tendency to straighten the line and to shorten it by cutting off some of the great curve through Mexico and Texas. First a short flight across the western part of the Gulf of Mexico was probably essayed. Proving successful, this was followed by flight lines that moved farther east, until finally the roundabout curve through Texas was entirely discarded and the flight made directly across the Gulf to southern Louisiana.

As the great areas in Canada were gradually added to the birds' domain, other important factors arose, the chief being the attractiveness of the vast stretches of coast and plain of the Labrador Peninsula, which in fall offered a bountiful store of berries. The fall route therefore worked eastward to the Gulf of St. Lawrence, thence southwest along the coast to Florida and across the Gulf of Mexico to the Central American mainland. A series of shortening flights followed to take out the great curve of the Atlantic coast. A relatively short ocean flight was probably attempted, say from Cape Cod to the Bahama Islands, Cuba, and Jamaica, followed eventually by the long direct oceanic route as it is now known.

As the Labrador Peninsula in spring is bound by frost and shrouded in fog while the season advances rapidly through the interior, the oceanic route proved useful only in fall, and the spring flight continued through the Mississippi Valley. This outline, although entirely hypothetical, gives a probable and fairly plausible explanation of the origin of this wonderful route, particularly when it is remembered that migration routes as now known are evolutions--age-long modifications of other routes.

The evolution of the migration route of the Pacific golden plover may be explained in a similar fashion. At first the route probably followed the Asiatic coast, through the Malay Peninsula and Oceania, thence east in a great curve to the Low Archipelago, with individuals and flocks dropping out to winter at many points along the way. The Siberian birds probably continue to follow this ancient highway, but those nesting in Alaska began a long evolutionary series of flights that cut down the length of their journey by shortening the curve, until finally the transoceanic route of the present day was developed.

This theory of the evolution of migration routes has been questioned by some ornithologists on the ground that it implies the possession in some degree of reasoning powers such as would be used by human beings. This opposition suggests that changes in migration routes might develop suddenly following mass survival of birds that were driven over the new route by a storm on some specific occasion. In the language of genetics, the new route would be, in effect, a mutation, rather than the result of an accumulation of infinitesimal variations. There is some evidence in support of this opposing theory. For example, information from the Hudson's Bay Co. post at Great Whale River, on the southeastern coast of Hudson Bay, indicates that in 1884 the snow geese suddenly changed their route from the eastern to the western coasts of Hudson and James Bays. According to one report, this change was caused by strong winds from the south which caught the birds in their fall migration and caused them to cross the entrance of James Bay from Cape Jones to the western side; the route thus reportedly forced upon them was then used in succeeding years.

_Vertical Migration_

In the effort to find winter quarters furnishing satisfactory living conditions, many North American birds fly hundreds of miles across land and sea. Others, however, are able to attain their objective merely by moving down the sides of a mountain. In such cases a few hundred feet of altitude correspond to hundreds of miles of latitude. Movements of this kind, known as "vertical migrations," are found wherever there are large mountain ranges. In the Rocky Mountain region they are particularly notable, as chickadees, rosy finches, juncos, pine grosbeaks, and some other species that nest in the Alpine Zone move down to the lower levels to spend the winter. It has been noted that such species as Williamson's sapsucker and the western wood pewee, which nest in the higher mountains, move down to the lower regions in August following the breeding season. There is a distinct tendency among the young of mountain-breeding birds to work down to the lower levels as soon as the nesting season is over. The sudden increases among birds in the edges of the foothills are particularly noticeable when cold spells with snow or frost occur at the higher altitudes.

Some species that normally breed in the Hudsonian or Arctic Zones find suitable breeding areas on the higher levels of the mountains, as for example the pipit, or titlark, which breeds on the tundra of Alaska and northern Canada and also south as far as Colorado on the summits of many peaks in the Rocky Mountains. On the other hand, a few species, as the Clark's crow, or nutcracker, nest at relatively low altitudes in the mountains and as the summer advances move higher up, thus performing a vertical migration that in a sense is comparable with the post-breeding movements of herons on the Atlantic coast. These illustrations show that the length of a migration route may depend upon factors other than latitude.

_Vagrant Migration_

The most striking feature of the migrations of some of the herons is a northward movement after the nesting season. The young of some species commonly wander late in the summer and in fall, sometimes traveling several hundred miles north of the district in which they were hatched. The little blue heron breeds commonly north to South Carolina, and by the last of July the young birds begin to appear along the Potomac, Patuxent, and Susquehanna Rivers, tributary to Chesapeake Bay. Although almost all are immature individuals, as shown by their white plumage, an occasional adult may be noted. With them come egrets and snowy herons and on occasion all three species will travel in the East as far north as New England, and in the Mississippi Valley to southeastern Kansas and Illinois. In September most of them disappear, probably returning south by the same route.

The black-crowned night heron has similar wandering habits, and young birds banded in a large colony at Barnstable, Mass., have been recaptured the same season north to Maine and Quebec and west to New York. This habit seems to be shared by some of the gulls also, although here the evidence is not so conclusive. Herring gulls banded as chicks at colonies in the Great Lakes have scattered in all directions after the breeding season, some having been recovered well north in Canada.

These movements may be considered as migration governed only by the availability of food, and they are counteracted in fall by a directive migratory impulse that carries back to their normal winter homes in the south those birds that after the nesting period attained more northern latitudes. They are not to be compared with the great invasions of certain birds from the North. Classic examples of the latter in the eastern part of the country are the periodic flights of crossbills. Sometimes these migrations will extend well south into the Carolinian Zone.

Snowy owls are noted for occasional invasions that have been correlated with the periodic maximum of Arctic foxes and the lemming cycle in the north. According to Gross (1947) 24 major invasions occurred between 1833 and 1945. The interval between these varied from 2 to 14 years, but nearly half (11) were at intervals of 4 years. A great flight occurred in the winter of 1926-27 when more than 1,000 records were received from New England alone, but the largest on record was in 1945-46 when the "Snowy Owl Committee" of the American Ornithologists' Union received reports of 13,502 birds, of which 4,443 were reported as killed. It extended over the entire width of the continent from Washington and British Columbia to the Atlantic coast and south to Nebraska, Illinois, Indiana, Pennsylvania, and Maryland. One was taken as far south as South Carolina.

In the Rocky Mountain region great flights of the beautiful Bohemian waxwing are occasionally recorded. The greatest invasion in the history of Colorado ornithology occurred in February 1917, at which time the writer estimated that at least 10,000 were within the corporate limits of the city of Denver. The last previous occurrence of the species in large numbers in that section was in 1908.

Evening grosbeaks likewise are given to performing more or less wandering journeys, and curiously enough, in addition to occasional trips south of their regular range, they travel east and west, sometimes covering long distances. For example, grosbeaks banded at Sault Ste. Marie, Mich., have been recaptured on Cape Cod, Mass., and in the following season have been retrapped at the banding station. Banding records demonstrate that this east-and-west trip across the northeastern part of the country is sometimes made also by purple finches.

_Perils of Migration_

The period of migration is a season full of peril for birds. Untold thousands of the smaller migrants are destroyed each year by storms and through attacks of predatory birds, mammals, and reptiles. If each pair of adult birds should succeed in raising two fledglings to maturity, the population of migratory birds would have a potential annual increase of 100 percent and the world would soon be heavily overpopulated with them. Since there is no such increase it is evident that the annual mortality from natural causes is heavy enough to keep it in check.

=Storms=

Of the various factors limiting the abundance of birds, particularly the smaller species, storms are the most potent. Special sufferers are those birds that in crossing broad stretches of water are forced down by a storm within reach of the waves. Such a catastrophe was once seen from the deck of a vessel in the Gulf of Mexico, 30 miles off the mouth of the Mississippi River. Great numbers of migrating birds, chiefly warblers, had accomplished nearly 95 percent of their long flight and were nearing land, when, caught by a norther against which they were unable to contend, hundreds were forced into the waters of the Gulf and drowned. On another occasion, on Lake Michigan, a severe storm came up at a time when large numbers of migratory birds were crossing and forced numerous victims into the waves. During the fall migration of 1906, when thousands of birds were crossing Lake Huron, a sudden drop in temperature accompanied by a heavy snowfall resulted in the death of incredible numbers. Literally thousands were forced into the water and subsequently cast up along the beaches, where in places their bodies were piled in windrows. On one section of the beach the dead birds were estimated at 1,000 per mile, and at another point at 5 times that number. Most of them were species that rank among our most desirable birds as destroyers of insects and weed seeds, including slate-colored juncos, tree sparrows, white-throated sparrows, swamp sparrows, winter wrens, and golden-crowned kinglets, together with many brown creepers, hermit thrushes, warblers, vireos, and others.

Of all species of North American birds, the Lapland longspur seems to be the most frequent victim of mass destruction from storms. These birds sometimes congregate in enormous numbers where grass or weed seed is abundant. Almost every winter brings in reports of their death by thousands somewhere in the Middle West. While migrating northward at night they have encountered blinding storms of wet, clinging snow, which have so bewildered them that they have flown into various obstructions, or have sunk to the ground and perished of exposure and exhaustion. In 1907 an experienced ornithologist estimated that 750,000 longspurs were lying dead on the ice of 2 lakes in Minnesota, each about 1 square mile in extent, and dead birds were reported in greater or less abundance on this occasion over an area of more than 1,500 square miles. The heaviest mortality occurred in towns, where, bewildered by the darkness and the heavy falling snow, some of the birds congregating in great numbers flew against various obstacles and were killed or stunned, while many others fell to the ground exhausted. Similar catastrophes have been reported from eastern Colorado, Nebraska, and North Dakota.

During the early part of June 1927, a hailstorm of exceptional severity in and around Denver, Colo., killed large numbers of robins, meadow larks, sparrows, and others. The lawns of parks were strewn with the bodies of these birds, and many lay dead in their nests where they were covering their eggs or young when the storm broke.

=Aerial obstructions=

Lighthouses, lightships, tall bridges, piers, monuments, and other aerial obstructions have been responsible for a tremendous destruction of migratory birds. Beams of the lanterns at light stations have a powerful attraction for nocturnal travelers of the air that may be likened to the fascination for lights that also is shown by many insects, particularly night-flying moths. The attraction is not so potent in clear weather, but when the atmosphere is moisture laden, as in a heavy fog, the rays have a dazzling effect that lures the birds to their death. They may fly straight up the beam and dash themselves headlong against the glass, or they may keep fluttering around the source of the light until exhausted, and then drop to the rocks or waves below. The fixed, white, stationary light located 180 feet above sea level at Ponce de Leon Inlet (formerly Mosquito Inlet), Fla., has caused great destruction of bird life even though the lens is shielded by wire netting. On one occasion an observer gathered up a bushel-basketful of warblers, sparrows, and other small passerine birds that had struck during the night. The birds apparently beat themselves to death against the wire or fell exhausted to the concrete pavement below, frequently to be destroyed there by cats or skunks. Two other lighthouses at the southern end of Florida, Sombrero Key and Fowey Rocks, have been the cause of a great number of bird tragedies, while heavy mortality has been noted also at some of the lights on the Great Lakes and on the coast of Quebec. It is the fixed white lights that cause such disasters to birds, as the stations equipped with flashing or red lights do not present such strong attractions. That it is not a mere case of geographical location has been demonstrated, for it is observed that when fixed white lights have been changed to red or flashing lights, the migrating birds are no longer endangered. At some of the light stations in England and elsewhere, shelves and perches have been placed below the lanterns to afford places where birds can rest until they have overcome their bewilderment.

For many years at the National Capital, the Washington Monument, although unilluminated, caused the destruction of large numbers of small birds, due apparently to their inability to see this obstacle in their path, towering more than 555 feet into the air. One morning in the spring of 1902 the bodies of nearly 150 warblers, sparrows, and other birds were found about its base. Then, as the illumination of the city was improved and the Monument became more visible at night, the loss became steadily less, until by 1920 only a few birds would be killed during an entire migration. On November 11, 1931, however, as part of the Armistice Day celebration, batteries of brilliant floodlights grouped on all four sides about the base of the Monument were added to the two searchlights already trained on the apex, so that the lighted shaft probably corresponds in brilliancy to a very low magnitude lighthouse lantern. Airplane pilots have ventured opinions that on a clear night it could be seen for 40 miles. It is certain that there is an extensive area of illumination, and on dark nights, when there are gusty, northerly winds and the nocturnal travelers seem to fly at lower altitudes, many of them are attracted to the Monument as to a lighthouse beacon. As they approach from the north a last-minute attempt to avoid it causes them to veer off to the east or the west where they are literally sucked in and dashed against the southern face of the shaft. During the fall migration of 1932 more than 500 warblers, vireos, thrushes, kinglets, sparrows, and others were killed. Since that year the mortality has been less, but the Monument at times remains a serious menace to birds during migration and some are killed nearly every fall.

When the torch on the Statue of Liberty in New York Harbor was kept lighted, it caused an enormous destruction of bird life, tabulations showing as many as 700 birds killed in a single month.

In September 1948, bird students were startled by news of the wholesale destruction of Maryland yellowthroats, redstarts, ovenbirds, and others that were dashed against the 1,250-foot high Empire State Building in New York City, the 491-foot high Philadelphia Saving Fund Society Building in Philadelphia, and the 450-foot high WBAL radio tower in Baltimore. In New York the birds continued to crash into the Empire State Building over a 6-hour period and their bodies were scattered over a four-block area. The mortality was so heavy in Philadelphia that it was impossible to use the sidewalk below the sky-scraper until the birds had been gathered. A study of the weather conditions prevailing at this time in the Atlantic coastal region suggests the probable cause of this catastrophe. By early morning on September 11 a mass of cold, southward-flowing air had just reached New York City where it was forcing upward and was being overridden by a mass of warm, northward-flowing air. Presumably the migrants were riding the upper levels of the southbound current which, in the contact zone with the northbound current, was being deflected earthward, thus causing the birds to fly lower and lower until they were below the tower of the Empire State Building. Clouds and gusty winds in the zone of contact between the two air masses reduced visibility and disrupted avian navigation with the result that the confused travelers crashed into the stone and steel obstruction. As the cold air mass continued to move southward, the situation was repeated at Philadelphia and at Baltimore.

=Exhaustion=

Although it would seem that the exertion incident to the long flights of many species of migratory birds would result in their arrival at their destination in a state bordering on exhaustion, this is contrary to the truth. Both the soaring and the sailing of birds show them to be proficient in the use of factors employed in aerial transportation that only recently have become understood and imitated by aeronautical engineers. The use of ascending currents of air, employed by all soaring birds, and easily demonstrated by observing the gulls that glide hour after hour along the windward side of a ship, are now utilized by man in his operation of gliders. Moreover, the whole structure of a bird renders it the most perfect machine for extensive flight that the world has ever known. Hollow, air-filled bones, making an ideal combination of strength and lightness, and the lightest and toughest material possible for flight in the form of feathers, combine to produce a perfect flying machine. Mere consideration of a bird's economy of fuel or energy also is enlightening. The golden plover, traveling over the oceanic route, makes the entire distance of 2,400 miles from Nova Scotia to South America without stop, probably requiring about 48 hours of continuous flight. This is accomplished with the consumption of less than 2 ounces of fuel in the form of body fat. To be as economical in operation, a 1,000-pound airplane would consume in a 20-mile flight not the gallon of fuel usually required, but only a single pint.

The sora, or Carolina rail, which is such a notoriously weak flyer that at least one writer was led to infer that most of its migration was made on foot, has one of the longest migration routes of any member of the family, and easily crosses the wide reaches of the Caribbean Sea. The tiny ruby-throated hummingbird crosses the Gulf of Mexico in a single flight of more than 500 miles.

While birds that have recently arrived from a protracted flight over land or sea sometimes show evidences of being tired--as, for example, pintail ducks that have flown from the North American mainland to the Hawaiian Islands--their condition is far from being a state of exhaustion. With a few hours' rest and a crop well filled with proper food, most birds exhibit eagerness to resume their journey. The popular notion that birds find the long ocean flights excessively wearisome and that they sink exhausted when terra firma is reached, generally does not agree with the facts. The truth lies in the opposite direction, as even small land birds are so little averse to ocean voyages that they not only cross the Gulf of Mexico at its widest point, but may even pass without pause over the low, swampy coastal plain to the higher regions beyond. Under favorable conditions birds can fly when, where, and how they please. Consequently the distance covered in a single flight is governed chiefly by the food supply. Exhaustion, except as the result of unusual factors, cannot be said to be an important peril of migration.

_Influence of the Weather on Migration_

The state of the weather at any point has little if anything to do with the time of arrival of migratory birds. This is contrary to the belief of observers who have thought that they could foretell the appearance of various species by a study of the weather conditions. Though the insistent crescendo note of the ovenbird is ordinarily associated with the full verdure of May woods, this bird has been known to reach its breeding grounds in a snowstorm and the records of its arrival in southern Minnesota show a temperature variation from near freezing to full summer warmth. Temperatures at arrival of several other common birds vary from 14 degrees between highest and lowest temperatures to 37 degrees, the average variation being about 24 degrees.

It should be remembered that North American species spending the winter months in tropical latitudes experience no marked changes in climatic conditions from November to March or April, yet frequently they will start the northward movement in January or February. This is in obedience to physiological promptings and has no relation to the prevailing weather conditions. For migratory birds the winter season is a period of rest, a time when they have no cares other than those associated with the daily search for food or escape from their natural enemies. Their migrations, however, are a vital part of their life cycles, which have become so well adjusted that the seasons of travel correspond in general with the major seasonal changes on their breeding grounds. With the approach of spring, therefore, the reproductive impulse awakens, and each individual bird is irresistibly impelled to start the journey that ends in its summer home.

In other words, the evidence indicates that the urge to migrate is so ingrained that each species moves north in spring when the average weather that will be encountered is not unendurable. The word "average" must be emphasized since it appears obvious that the migrations of birds have so evolved that in general they synchronize with average climatic conditions. The hardy species travel early, fearless of the blasts of retreating winter, while the more delicate kinds come later when there is less danger of encountering prolonged periods of inclement weather. Some of the hardy birds pause in favorable areas and allow the spring season to advance. Then, by rapid travel they again overtake it, or, as sometimes happens, they actually outstrip it. Occasionally this results in some hardship, and rarely in the destruction of large numbers of individuals. Cases are known where early migrating bluebirds have been overwhelmed by late winter storms. Nevertheless, unless such climatic conditions are prolonged, no serious effect on the species is noted. The soundness of the bird's instincts is evidenced by the fact that natural catastrophes, great though they may be, do not permanently diminish the avian populations.

As has been pointed out, the advance of average temperature lines, known as isotherms, is found to correspond closely with the northward movements of certain species. For example, the northward travels of the Canada goose are found to coincide with the advance of the isotherm of 35° F. (fig. 4).

The spring flight of migrants, if interrupted for any reason, is resumed when weather conditions again become favorable, and it is probable that all instances of arrival of birds in stormy weather can be explained on the theory that the flight was begun while the weather was auspicious. The state of the weather when a flight starts at any southern point, the relation of that place to the average position of the bird under normal weather conditions on that date, and the average rate of migratory flight, are data basic to any reasonably accurate prediction of the time arrival may be expected in northern areas.

Head winds are as unfavorable to migration as is rain or snow, as they greatly increase the labor of flight and cut down the speed of cross-country travel. If such winds have a particularly high velocity they may force down the weaker travelers, and when this happens over water areas, large numbers of birds are lost. Even strong winds that blow in the direction of aerial travel are unfavorable for the birds, as they interfere with their balance and disarrange their feathers. Moderate tail winds and cross or quartering breezes appear to offer the best conditions for the passage of the migrants.

_Problems of Migration_

=Banding studies=

The study of living birds by the banding method, whereby great numbers of individuals are marked with numbered aluminum leg rings, has come to be recognized as a most accurate means of ornithological research. Since 1920, banding work in North America has been under the direction of the Fish and Wildlife Service in cooperation with the Dominion Wildlife Service of Canada. Every year voluntary cooperators, working under permit, place bands on thousands of birds, game and nongame, large and small, migratory and nonmigratory, each band carrying a serial number and the legend, NOTIFY FISH AND WILDLIFE SERVICE, WASHINGTON, D. C., or on the smaller sizes an abbreviation thereof. When a banded bird is reported from a second locality, a definite fact relative to its movements becomes known, and a study of many cases of this nature develops more and more complete knowledge of the details of migration.

The records of banded birds are also yielding other pertinent information relative to their migrations, such as the exact dates of arrival and departure of individuals, the length of time that different birds pause on their migratory journeys to feed and rest, the relation between weather conditions and the starting times for migration, the rates of travel of individual birds, the degree of regularity with which birds return to the exact summer or winter quarters used in former years, and many other details that could be learned in no other manner. Banding stations that are operated systematically throughout the year, therefore, are supplying much information concerning the movements of migratory birds that heretofore could only be surmised. (See Appendix II, p. 92 for instructions on reporting the recovery of banded birds.)

=Movements of residents=

Typical migration consists of definite movements that are repeated regularly year after year, and it is to these that the term is generally restricted. It is desirable, however, if only for purposes of comparison, that some account be taken of the movements of some other birds, which, while not typical, do possess some of the characteristics of true migration. Data on this subject are being collected through bird banding.

There are several species that are customarily grouped under the heading "permanent residents," the term implying that these birds do not travel but remain throughout the year in one locality. Among these are the cardinal, the tufted titmouse, the wrentit, the Carolina wren, the house finch, the bobwhite, the California quail, and the ruffed grouse. Each species may be present constantly throughout the year, although in the northern part of the range there is probably a slight withdrawal of the breeding birds in winter. The individuals to be seen at that season, therefore, may not always be the same as those observed during the summer. It is certain, however, that these species do not regularly perform extensive journeys.

While the blue jay is disposed to be secretive, it is such a showy and noisy bird that it is not likely to escape notice. In the vicinity of Washington, D. C., as in many other places, it is present the year round, but at the end of September or early in October when the weather is becoming cooler, troops of jays are sometimes seen working southward through the trees. A corresponding northward movement occurs again in May. This is unquestionably a migration to and from some winter range, but its extent or significance is not now known. Some light is being shed on the matter, however, through the records of banded birds, and these eventually will fill in a more perfect picture of the movements of this species. One jay, banded on September 14, 1923, at Waukegan, Ill., was killed at Peruque, Mo., on November 15 of the same year; another, banded at Winnetka, Ill., on June 16, 1925, was retaken at Sulphur Rock, Ark., the following December 10; a third, banded on May 6, 1925, at Whitten, Iowa, was recaptured at Decatur, Ark., on January 22, 1926. These three birds unquestionably had made a flight that had every appearance of being a true migration to winter quarters in Missouri and Arkansas.

The black-capped chickadee is apparently resident in many places, but occasionally in winter it invades the range of the southern Carolina chickadee, and in northern Canada it is regularly a migrant.

In the coastal plain between Washington, D. C., and the Atlantic Ocean, the white-breasted nuthatch is usually absent during the summer, nesting at that season in the higher, or piedmont, country. Late in fall, however, it appears in fair abundance in the wooded bottoms, remaining at the lower levels until the following March or April.

Some birds, including the screech owl, bobwhite, Carolina wren, and mockingbird, seem to be actually sedentary, but even these are sometimes given to post-breeding wanderings. Ordinarily bobwhites that are marked with numbered bands are seldom retaken far from the area where banded, but sometimes they will travel 10 miles or more. A screech owl banded at Glenwood, Minn., in March, was recovered the following December at Emmetsburg, Iowa, 180 miles south. Such flights, however, are probably more in the nature of a search for new feeding areas, or to escape from a winged enemy, than a true migratory journey.

=Migration of the white-throated sparrow=