PROBLEMS OF MEDITERRANEAN‐SAHARAN MIGRATION

Summary It is estimated that birds travelling to winter south of the Sahara must on average enter Africa at the rate of 250,000 per mile of longitude, which over a two‐month period gives an average daily entry of about 4,000 birds per mile. Only a small fraction of these come within the range of observation. There are reasons to suppose that the main directions may be southwest in autumn and northeast in spring, involving somewhat diagonal crossings of the desert, rather than directly north and south. The geography, ecology and winds of the Mediterranean and Sahara are described. The average width of the Sahara from north to south is at least 900 miles, but owing to the aridity on the southern edge in spring (mitigated in autumn) many birds probably start their flight up to 1,200 miles south of the Mediterranean. There are indications that migrants do not concentrate on oases. Wind data are given for altitudes up to 3,000 m. (10,000 ft.), which on the whole are much more favourable to birds in autumn than in spring, but with less difference in the sectors Algeria‐Tripoli than elsewhere and with the proviso that in spring at 2,000 m. upwards the winds are favourable to diagonal flight, northeastwards. Hypothetical flight‐performances are estimated and discussed; and it is concluded that ability to fly some 50–60 hours without refuelling is essential for most migrants in spring. The physiological implications are considered. If as much pre‐migration fat is put on in autumn as seems to be needed in spring, then many birds should be able to fly non‐stop from Europe to the tropics, some 1,300‐1,500 miles. Ringing in Tunisia and subsequent recoveries there have provided evidence of accurate navigation directed to a point half‐way through the birds' journeys. The evidence for general abundance of birds in spring and autumn is assembled sector by sector. In the northern deserts of Algeria, in Tunisia and on the coasts of Tripoli and Cyrenaica far more migrants are seen in spring than in autumn, while further east and also in Morocco, the disparity seems to be less. It is concluded that, at least in the central sectors named, a much larger proportion of birds makes a more or less continuous flight, without coming under observation, in autumn than in spring. The information regarding the incidence of 57 species of trans‐Saharan migrants is summarized and discussed for each. A number of species have almost entirely escaped detection, especially in autumn and especially in the eastern half of the area. Most species of migrant are shown to travel over various sectors of the Sahara, irrespective of opportunities to refuel, at both seasons and it seems that many birds fly non‐stop from Europe to the tropics in autumn. Indications that some species or populations travel further east at one season than at another are discussed. “Migration divides” which have been detected in parts of Europe are as a rule not reflected by differences in abundance on the south side of the Mediterranean.