The relative importance and biological significance of non-stop and intermittent flight strategies in bird migrants crossing the Sahara

IX Summary
The aim of this thesis was to investigate the flight and resting strategies of birds crossing the Sahara on migration. In order to overcome the severe limitations of ground observations, we chose radar to obtain continuous data of the temporal and spatial distribution of bird migration and to answer the core question whether migrants cross the largest desert by a long non-stop flight or rather by intermittent flying and resting. Study site and methods The Swiss Ornithological Institute carried out extensive radar and field studies at three different study sites in Mauritania, covering three migration seasons in 2003 and 2004. One station (20°56’ N; 11°35’ W), close to the oasis Ouadâne and about 500 km east of the Atlantic coast, was run permanently for about two moths in spring and autumn 2003 and in spring 2004. It provided continuous data on the course of bird migration across this part of the Sahara. The two other radar sites were operated about 300 km further to the East in the plain sand desert close to the border of Mali. At Mohammed Lemna (18°35’ N; 08°38’ W) radar data was collected for eleven days in September 2003. At Bîr Amrâne (22°47’ N; 08°43’ W), located about 450 km further to the North of Mohammed Lemna, the observations covered about three weeks in April 2004 (chapter I). Quantification of bird migration by means of radar is still a difficult task, in spite of almost sixty years of growing experience in radar ornithology. We present for the first time a method how radar data can be analysed to achieve reliable estimates of the number of migrating birds in a given air space. To reach this goal three basic steps of quantification must be considered: (1) echo detection, (2) echo identification and (3) quantification. First, the radar must be calibrated to allow the calculation of standardized echo sizes that are independent of distance. Then a distance-dependent threshold to reduce the detection of small targets at close range has to be applied. After this basic step, the remaining echoes must be assigned to bird or insect targets, eventually eliminating the insects. In our case, we were also able to determine the birds’ body axis relative to the radar (aspect) and to differentiate detection probabilities according to echo sizes. For the final quantification, we estimated the surveyed air space with respect to the specific detection range per echo size class. Using this method we calculated the temporal and spatial distribution of bird migration crossing the western Sahara (chapter II). Composition of daytime migration Billions of songbirds breeding in the western Palaearctic cross the Sahara desert twice a year. While crossing temperate latitudes most of them use an intermittent flight strategy, i.e. fly at night and rest or feed during the day. As resting and feeding in the Sahara is much less rewarding than in Europe, we investigated whether in the Sahara migration during daytime comprises the normal diurnal migrant species or the nocturnal ones prolonging their flight into the day. At the radar station Ouadâne, spring passage of passerine migrants showed high variation from night to night, but in general passerine densities decreased towards sunrise. However, under favourable wind conditions passerine migration continued far into the day, mostly at high altitudes (around 2000 m above ground level). These relatively high densities of passerines flying during the day were in contrast to very low proportions of real diurnal migrants detected on daily transect counts in the study area. For nocturnal migrants, however, the transect counts showed high correlation IX-1