A Continentwide View of Bird Migration on Four Nights in October

Abstract

IN 1952, between twilight on 1 October and dawn on 5 October, observers watching through telescopes witnessed the passage of 35,407 presump-tive migrants silhouetted against the moon. No less than 1,391 bird students and astronomers joined in the organized effort that produced this total. They manned 265 observation stations at 235 named locations representing three of the provinces of Canada and every state of the United States except Alaska, Hawaii, Maine, Nevada, and Utah. Figure 1 and its accompanying key identify the locations. The records produced by this project offered a basis for constructing continentwide synoptic views of nocturnal migration over North America. But the difficulties were formidable. We have spent more than a decade in overcoming them. TIlE METIIOD OF PORTRAYING MIGRATION Lunar observations provide information regarding two features of the migration over the vicinity of the observing point-the approximate amount of migration and its directional trends. Neither the quantity of migration nor its directions, however, are immediately discernible from the raw data recorded in the field. Each must be decoded from these data by a mathematical procedure that compensates for the changing effects exerted on observations by the constantly moving moon. By 1952, mathematical means of correcting for changes in the size of the effective observation space due to the movement of the moon and of calculating flight directions from the apparent pathways of birds as seen against the moon had already been devised (Rense, 1946 and 1950; Low-ery, 1951). But the procedures were extremely cumbersome and time-consuming. To cope with the data amassed in 1952, we developed a new computation system based on the principle of the old one but incorporating a large measure of precalculation. The mathematical results express the quantity of migration in terms of the theoretical number of passing birds per mile of front per hour. Our earlier papers referred to this theoretical number as a /fight density. Because density primarily denotes the quantity of objects in space alone and because "flight density" in our sense relates to numbers in space and time combined, we now prefer the term migration tra/fi'c rate. The migration traffic rates on which the present paper is based are stated in detail in a supplemental publication (Newman and Lowery, 1964). 547