Results and material from Sweden have contributed little to the German papers. Although of long standing, the study of Dunlin migration in the Swedish Baltic area has long been hampered by incomplete data, caused by the abridged catching season at two important stations (Ottenby, Falsterbo; Pettersson 1994, Roos 1984). This incompleteness of data adds to the burden of Dierschke and Rösner (but in particular it embarasses the conclusions concerning autumn in Goede, Nieboer & Zegers 1990), and it can only partly be remedied with material from the Bay of Gdansk, Reda mouth (Gdansk area, Poland) and Langenwerder (Bay of Wismar, Germany); only at Langenwerder has there been some focus on late material. Meltofte 1993 stands closer to the Baltic than the above authors, he has a more profound understanding of "Baltic" issues, and his views are also to some extent rejected by Rösner 1997. The German and Dutch authors do not have the full picture of Dunlin migration in the Baltic; not of the extent of adult migration, not of the extent of moult, and not of the relative adult-juvenile patterns in the Baltic area, and this lack of information combines to "colour" in particular Rösner's conclusions with a certain Waddensea-induced indeterminacy. What appears as stochastic confusion or at best trial-and-error progress (in the case of Rösner acting as an invitation to stochastic modelling) in the crowded Waddensea is reduced to apparent goal-orientation and strict regulation of the spatial distribution of adults and juveniles when studied from the vantage point of the less crowded Baltic area.

In the papers linked to this introduction new aspects of Dunlin phenology, biometry and moult, collected in the Sound area, S. Sweden and based on all-the-year-round catches, are presented. The staging of juveniles in the Sound area in October - November, followed by departure for unknown goal-areas is novelty, as is the encounter of juveniles and adults belonging to the wintering population and the morphological cline of autumn juveniles. The central aim is to give a complete picture of phenology (in particular the phenology of adult migration), biometry (actual span of measurements and sex differences) and the extent of moult in the Baltic area. These sections have bearing on moult papers by Greenwood 1983, 1986, Gromadzka 1986 and Holmgren et al. 1993a, 1993b, (as well as on views on moult held by Rösner 1990, 1997 with Holmgren as intermediary) and on discussions of biometrical issues by Brenning 1987, Goede, Nieboer & Zegers 1990, Greenwood 1979, 1986, Meltofte 1993, Stiefel & Scheufler 1989 and Engelmoor & Roselaar 1998. A preliminary algorithm for separating non-normal bill-length distributions is presented in "Bill length distributions...", it will be revised by degrees, with better, more reliable material. More details about the points of criticism are given in the introductions of the specific papers.

The maximum tidal amplitude in the Belt Sea area is 0.2 - 0.3 m, a little more where the Danish straits enter Kattegat in the north (i.e. less than 5 % of amplitudes on the Wash). On 9.4.05, with new moon, air pressure rising from 1005 to 1010 hPa and wind c8 m/s NW, the water level at Drogden (in the Sound, close to Copenhagen) oscillated from - 27 cm (02h) over -13 cm (07h) and back to - 28 cm (12h), that is an amplitude of 15 cm under these particular conditions. The most pronounced tidal effect coincides with the full moon, causing the water level to rise 10 - 15 cm twice a day; otherwise tide is overshadowed by changes induced by air-pressure and wind. In the Sound the latter may exceed 1 m during storm-floods in late autumn and winter, and in later years in spring as well. This state prevails at least from Germany (Rügen area) to Sweden (Falsterbo Peninsula) and north Sealand (Isefjord); the important feeding-grounds visited by waders in the marine environment are wind flats. This fact creates a particular bond between the Baltic and the Mediterranean and the Black Sea area, where wind flats are also the main feeding-grounds of resting Dunlin (van der Have & van den Berk 1994). In the south Kattegat area the situation is more balanced; in calm weather tide is a notable and sustained phenomenon, but with winds from the west or north its effects subside even here, and foraging waders may be forced to enter the flooded fjords and seek shelter on meadows surrounding them.

The average salinity in the south Baltic east of the Darß threshold (Fig. 1 in "Wintering and spring staging Dunlin...") is 7 - 10 ‰. This water moves north as a surface current through the Belts and the Sound and meets with a southbound bottom current of marine salinity (30 - 34 ‰). At the "Belt front", and to some extent in the Sound, salt and brackish water intermingle, with a resulting salinity of 15 - 20 ‰. The latter figure is also valid for surface water over great parts of Kattegat. This pattern in turn is influenced by wind; with a wind shift from south to north the salinity in the Sound may increase from 10 to 15 ‰ practically overnight, creating stress on all water-living organisms. In ice winters the freezing date of coastal waters is influenced by salinity and currents; the south end of Kattegat will offer open water (and a more diverse and rich supply of benthic organisms) one or two weeks longer than the German and Polish shores 200 kms further to the south.
The Falsterbo peninsula is situated where the Sound enters the Baltic. It is framed by a system of spits, including the islet of Måkläppen to the south (Fig. 1). Since the Sound constrains the water-masses moving through, water-levels at the south end will be particularly variable. With high water the spits get totally submerged, at the other extreme, with anticyclonic weather and winds from south or east, wind-flats will cover square miles. (The Bock and Bessin areas on Hiddensee/Rügen are equivalents; for an extensive study of windflat ecology in the Baltic see Dierschke & Helbig 1999, Dierschke et al. 1999a, Dierschke et al. 1999a). In addition Falsterbo is the only Swedish weather station with a mean February temperature in excess of 0° C. These qualities attract waterbirds attempting to spend the winter as far to the north and east as possible, and Curlew Numenius arquata, Iceland Redshank Tringa totanus robusta, Dunlin Calidris alpina and Shelduck Tadorna tadorna have been regular winter visitors to the area for at least three decades.

Fig. 1 The Falsterbo peninsula. Circled areas = sites where Dunlin have been caught in the present study. The Foteviken site is used mainly when all wind flats are flooded, V. Hake when wind flats have maximum extension.

On the peninsula proper migrating waders (in particular Dunlin) were ringed during summer months between 1964 and 1975 by Falsterbo Bird Observatory (Roos 1984). This catch never involved late autumn migrants or wintering birds. In the same area I began to ring late autumn and winter waders on a more regular basis in 1991, all parts of the peninsula being gradually involved. July / August adults were included from 1995. These catches were made during night-hours, employing regular wader-nets and tape recorders playing endless tapes as lures. (For details on the catching technique, see Busse 2000 (not active)). All birds were ringed and investigated in the field, on the wind flats, with forehead torches as light-source. When there is a wind, an exposed position of this kind may make weighing difficult, but as a matter of fact it makes catching difficult as well, so in most cases the biometry is complete. Much the same pattern was repeated in the years to come, with excursions to a few sites north of Falsterbo peninsula (Klagshamn, 55° 31' N, 12° 56' E) and April catches at Foteviken added in 1995. 122 birds were caught at Barsebäck (55° 45' N, 12° 54' E) in December 1995 and 1996 by Ulf Lundwall and Peter Olsson. From 1994 an orange colour-ring has been added above the metal ring on the right tarsus in winter birds (16.11 - 28.2), so far some 400 have been marked in this way. Numbers from July to February are given in Table I.

Most birds were measured, weighed and investigated for moult. Fat ranking is extremely difficult with forehead torches as light source; the light can not be properly directed when the down is blown aside, so all attempts at fat ranking in the dark have been abandoned. There is some daytime material, though, but still too small to allow safe conclusions. Up till and including April 1995 wing lengths were measured with straightened tip and flattened wing, but the lateral curve was not straightened by moving the tip and pressing near the base of primaries. This latter method in our case adds 2.5 mm to wing length, this is the constant mean difference between "before" and "after" measurements. Early measurements, when used, have been corrected by 2.5 mm. From May 1995 the maximum length technique (cf. Evans 1986) has been in use. Bill lengths were measured to the nearest 0.1 mm from tip of bill to feathers, weights with Pesola spring balances to the nearest 0.5 grams. Moult was recorded in tenths of full-grown feathers; this scale, widely used in Sweden, gives better resolution than the conventional Euring 5-grade scale (e.g. Ginn & Melville 1983). The age of birds is rendered by calendar years and denoted 1c, 2c, 2c+, 3c+ etc.