157 BEHAVIOURAL ASPECTS OF THE DEPARTURE OF WADERS BEFORE LONG-DISTANCE FLIGHTS: FLOCKING, VOCALIZATIONS, FLIGHT PATHS AND DIURNAL TIMING THEUNIS PIERSMAI,2,3, LEO ZWARTS3,4 & J. HENRICH BRUGGEMANN3,5 ABSTRACT The departure behaviour of waders was studied on the Bane d' Arguin in Mauritania in spring, and in The Netherlands in both spring and autumn. Most flocks consisted of 10-60 individuals, with the larger species occurring in larger flocks. When conspecifics were best 'available' (during peak migration and at high tides) starting flocks were largest. During depar- ture flocks sometimes joined others or split up, 'stabilizing' between 30-60 birds. Starting flocks usually assembled from clusters into V-formations or echelons. About 4% of the flocks contained more than one species. In more than 90% of the departing flocks birds vocalized intensely. With an optical range-finder we were able to follow flocks leaving the Bane d' Arguin to heights of 1.5 km, with no signs of flocks levelling off. During this ascent birds usually headed into the wind with climb rates ofO.55-0.69 m s·l. Except under the rare condition of tail winds, departing flocks showed no compen- sation for wind drift. A fixed orientation would make it easier for the birds to calculate wind drift effects between successive geographical fixes on which to base subsequent orientational adjustments. All species headed towards NNW, instead of NNE where the next stopover areas are found. This westerly orientation might represent an adaptation to correct for a continuous easterly drift at high altitude. Most flocks left before sunset, the departures showing a variable relationship to tide height. In the Dutch Wadden Sea, where cloud conditions vary, all 277 witnessed departures occurred with a visible sun. The diurnal timing of departures and the patterns of flocking can be understood as a result of the interactions between orientational and ener- getic advantages. IZoological Laboratory, University of Groningen, P.O. Box 14, 9750 AA Haren, The Netherlands; 2Netherlands Institute for Sea Research (NIOZ); 3Poundation Working Group for International Wader and Waterfowl Research (WIWO); 4Rijkswaterstaat Plevoland, Lelystad; 5Department of Marine Biology, University of Groningen. INTRODUCTION Migration can be 'in the air' even before waders fly up and leave an area. The increase in noisiness of the flocks as they assemble on their high tide roosts, and the occasional rounds of smaller parties up in the air and 'tornado-tumbling' down on the mudflats, indicate with certainty that something dramatic is about to happen: the departure of small and seemingly fragile flying creatures on some- times many a 1000 km long flight. There are sever- al questions that come to mind when one is left Ardea 78 (1990): 157-184 alone on a darkening Mauritanian saltflat or along a Dutch dike, Where do they go, how long will it take them, and will they make it? And: why go in flocks of this size and not alone, why vocalize so much, why fly in Vee's, and why leave at this time of the day? In spite of the fact that migratory departures of waders provide observant biologists with an evoca- tive experience, the behavioural phenomenon has received surprisingly scant attention in the liter- ature. With the notable exceptions of Lank (1983, 1989) and Alerstam et al. (1990), only incidental