lbis (2001) 143, 450-455

Detection of a new important staging and wintering area of the White Stork Ciconia ciconia by satellite tracking

PETER BERTHOLD’*, WILLEM VAN DEN BOSSCHE2, WOLFGANG FIEDLER’, CHRISTOPH KAATZ3, MICHAEL KAATZ3, YOSSl LESHEM4, EUGENIUSZ NOWAK5 & ULRICH QUERNER’

Research Centre for Ornithology of the Max Planck Society, Vogelwarte Radolfzell, 78315 Radolfzell, Germany ‘Belgische Natuur- en Vogelreservaten, BNVR-Birdlife Belgium, 1030 Brussels, Belgium

3Staatliche Vogelschutzwarte, Storchenhof Loburg, 39279 Loburg, Germany 4Dept. of Zoologx Tel Aviv Universitx 69978 Tel Aviv, Israel

5Langenbergsweg 77, 537 79 Bonn 2, Germany

Since 1991, a large-scale satellite traclung study of White Stork Ciconia ciconia has followed 75 individuals along the eastern migration route, which passes across Israel into eastern Africa. Twenty-six of these birds travelled at least as far as the Sudan. Fifteen (58%) did not migrate to the eastern Sudan - the primary winter quarters in Africa suggested by ring recoveries - but flew to western Sudan and Chad and in one case even through Cameroon and into Nigeria. The significance of this new, important staging region for White Storks should be investigated urgently, to establish the areas that are ecologically valuable for the conservation of this species.

The various European populations of the White Stork Ciconia ciconia are known to overwinter mainly in Africa and to reach this continent by routes that large- ly bypass the Mediterranean Sea. There are two such routes, separated by a ‘migration divide’: the western populations fly over the Iberian Peninsula and Gibraltar, primarily to northwestern Africa, whereas the eastern populations pass over Israel and the Sinai Peninsula into eastern Africa, some as far as the south- ern tip of the continent. For the eastern migrants, the Sudan is an important stopping place during winter. Some Storks rest there briefly, others use it as an inter- mediate destination for longer sojourns, and for still others it is the final winter quarters (Schuz et al. 1971). The recoveries made here in almost 100 years of White Stork ringing (Jespersen & T h i n g 1950) and many visual observations made on expeditions have suggested that the eastern Sudan is the most important region for White Storks.

In his review of White Stork migration, Schulz (1988) summarized the situation as follows: ‘In the short-grass savannah belt of the eastern Sudan ... the White Storks make a ... stopover, and with energy reserves replenished they then migrate further south, in order . . . to spend the winter months in eastern or

*Corresponding author Email berthold@vowa ornithol.mpg.de

southern Africa. The most important rest region extends . . . along the highway Wad Medani-Gedaref . . southeast of the capital Khartoum’. In contrast, obser- vations in the western Sudan are said to be only ‘relatively frequent’, and documentation in Chad shows that ‘the country does not lie within the regular overwintering region’ and any presumed members of eastern populations encountered there ‘must have departed from the actual migration route in Egypt or the Sudan, diverging to the west’. Schuz et al. (1971) and Nikolaus (1 989) also emphasize the dominant role of the eastern Sudan for White Storks, and Schuz and Bohringer (1950) even refer to birds documented in the western Sudan as having been ‘derailed’. However, the satellite traclung studies of the White Stork that we have been carrying out since 1991 show that, as a stag- ing area and intermediate destination of the White Stork, the western Sudan and Chad are at least equal in significance to the eastern Sudan. That finding is the subject of this report.

METHODS

Ring recoveries

White Storks are among the pioneer species of bird ringing. They have been ringed in large numbers since 1901 (Jespersen & T h i n g 1950, Schuz et al. 1971) and

@ 2001 British Ornithologists’ Union

Staging and wintering of White Stork 451

the recovery rate is high (c. 10%): out of 300 000 ringed White Storks, about 35 000 recoveries have been made. We have analysed the EURING (the con- federation of European Ringing Centres) databank to collate all White Stork ring recoveries. Here we report the 159 usable recoveries that are relevant to the geo- graphical region discussed below. These are recoveries for which the location is specified to a precision of at least 10 minutes of arc.

Satellite tracking data

From 1991 to 1999 we have tracked the migratory movements of 75 White Storks of the eastern popula- tions by satellite. The birds were equipped with transmitters in Germany (54), Poland (1 1) and Israel (lo), and 26 travelled into Africa a t least as far as the Sudan, where they rested. The data from these 26 birds - 19 adults and seven first-year birds - are includ- ed in this study. The satellite traclung was carried out by way of the ARGOS system in Toulouse. Basic descriptions of the satellite tracking of White Storks have been published, e.g. by Kaatz (1995) and Berthold et al. (1997). Regarding the present paper, the following additional information is significant: the 19 adults were caught and fitted with transmitters in Eastern Germany (eight, 1993-98), Poland (nine, 1996-98) and Israel (two, 1995 and 1996), the seven first-year birds all in Eastern Germany (1992-98).

The transmitters we used were obtained from the Japanese firm Toyocom in 1991, and from 1992 on from Microwave, USA; they became progressively lighter, weighing 90 g until 1994, 35-60 g from then until 1997, and 35 g from 1998 on. Those used until 1994 were mainly battery-powered, whereas solar energy was the predominant power source from 1995 to 1997 and has been used exclusively since the begin- ning of 1998. The transmitters were attached to the backs of the birds by means of lightweight (c. 20 g) Perlon harnesses, in such a way that the antenna pro- jected out of the back plumage for about 15-20 cm. So that there would be no sites of concentrated pressure or friction, the Perlon strands were coated in Teflon. The transmitters were adjusted so that we received locations daily from 21 birds and every two days from the remaining five.

Study area

The area within which the satellite tracked White Storks moved in the Sudan and Chad, and stayed for a t least 12 days, extended from 18 to 37"E and 9 to 17"N.

Therefore we selected this rectangular region for the comparison with ring recoveries (Fig. 1).

RESULTS

Ring recoveries versus satellite tracking data

In Figure 1 we show the locations of White Storks obtained by satellite tracking along with the ring recov- eries from the study area defined above. The distribution of the 159 recoveries on either side of the midline of the research area (which is at 27'30'E) is as follows: to the east, 139 (87%); to the west, 20 (13%). In contrast, only 11 (42%) of the satellite tracking loca- tions of stopover periods were east of the midline, and 15 (58%) were to the west. The difference is highly significant (x2 = 29.65, P < 0.0001).

Table 1 shows that the east-west distribution of the ring recoveries, with a clear bias towards the eastern half of the test area, is typical of all of the main regions of origin and ringing centres involved. Of the 26 tracked birds that reached the Sudan, 15 originated in Germany, nine in Poland and two in Israel, and ten of the birds from Germany plus five from Poland migrat- ed into the western half of the specified study area. Neither in the ring recoveries nor in the satellite track- ing locations were any differences found in the migratory and staging behaviour of adults and first-year birds in the area investigated.

Movements within Sudan and Chad, as far as Nigeria

Our results from satellite trachng enable us to docu- ment, for the first time, the complicated passage and

Table 1. Distribution of the 159 ring recoveries between the eastern and western parts of the study area, categorized by country of origin and ringing centre.

East Sudan Country I centre (east of 27'30'E) West Sudan Chad

Germany Helgoland Hiddensee Radolfzell

Hungary Latvia Lithuania Netherlands Poland Russia Yugoslavia (former)

39 8 2 12 1 49 3 1 6 1 1

2 3

2 1

26 1 1

@ 2001 British Ornithologists' Union, Ibis, 143, 450-455

452 Peter Berthold et al.

Figure 1. Location of the study region in Africa (inset), distribution of the staging areas of 26 White Storks Ciconia ciconia tracked by satellite in the study region in Africa (rectangular frame) in Chad and the Sudan (upper map) and distribution of 159 ring recoveries of White Storks in the same study region (lower map).

stopover patterns of White Storks in the Sudan and Chad. For this purpose we use the data from 20 of the 26 individuals included in this study, which are partic- ularly informative and typical (Fig. 2).

Most of the tracked storks, when passing through Egypt south of Aswan, did not follow the Nile as i t turned southwest but rather continued through the Nubian Desert, flying either south-southeast or south-

southwest. Both the eastern and the western migrants later rejoined the Nile, in a region approximately betLveen the localities Abu Hamed in the east and Dongola in the west. Thereafter, the eastern a n d west- ern migrants diverged progressively further. Birds aiming towards the eastern Sudan proceeded in the direction south-southeast to southeast, and after about 500-600 km reached their first African staging areas

0 2001 British Ornithologists' Union, /b/s. 143, 450-455

Staging and wintering of White Stork 453

Figure 2. Migratory routes of 20 White Storks tracked by satellite on their way from Egypt to staging areas in Chad and the Sudan. In these birds, transmitters remained intact so that all their movements during staging could be located. Dots: locations of individuals dur- ing migratory movements; dotted areas: numerous locations in narrowly circumscribed regions, which indicate stopover sites; a+: main staging areas in the eastern and western Sudan and Chad, respectively. Broken line: migratory route of an individual from Chad into Nigeria and back.

south of Khartoum (a). The storks with destinations in the western Sudan and in Chad at first continued to fly south-southwest, then southwest, until after about 1000 km they had reached the principal rest area in the western Sudan, in the Dafur region (b), and after a further 500 km, the principal resting regions in Chad, southwest of Abeche ( c ) . One of the storks tracked by us even migrated further west than Chad, as far as Nigeria. Having passed all the way through Chad, it crossed northern Cameroon on November 20, 1994 and reached eastern Nigeria (Fig. 2) on November 2 1, only to leave again on the same day. Then it flew back over northern Cameroon into Chad and settled in an area about 200 km southeast of Fort Lamy, where it remained into January 1995. This is the first demon- stration that storks of the eastern populations, initially following the eastern route, can reach the winter quar- ters of the western storks in Nigeria.

Tiventy of the tracked storks whose transmitters went on worlung departed from their resting regions in the Sudan and Chad between October and the end of November, thus identifying these areas as intermediate goals in accordance with previous ringing results, and continued to migrate, in some cases as far as South Africa (Berthold et al. 1997). In so doing, storks from the eastern Sudan simply carried on flying south, while those leaving Chad and the western Sudan more or less reversed their original migration direction and flew towards the east or southeast, so that in the southern Sudan they were reunited with their conspecifics from the eastern Sudan.

Of the 26 storks tracked into the study area in the Sudan and Chad defined above, we ceased to receive signals (as a result of death or transmitter failure) from five in the western half of the area (three adults, two juveniles) and only one (a juvenile) in the east. The

@ 2001 British Ornithologists' Union, Ibis, 143, 450-455

454 Peter Berthold et al.

remaining 20 birds subsequently left the study area and continued southwards, as described above, and ten of them were located later at sites ranging into south- ern Africa. Six of the 26 tracked storks were followed in two to three more migration periods, some with new transmitters. Five of these birds behaved the same on the second outward migration as on the first with respect to their choice of eastern or western interme- diate goals, whereas one thrice-tracked female rested in the eastern half of the study area on the first and third outward journey but in the western half on the second.

DISCUSSION

The comparison between the White Stork ring recov- eries, supplemented by visual observations, and the locations established by satellite tracking studies in the Sudan and Chad shows that the two methods - ringing and tracking - yield different distribution patterns: ring recoveries are chiefly concentrated in the eastern Sudan, while tracking data points to equally the west- ern Sudan and Chad. The differences might be due to a bias resulting from variation in the likelihood that ring recoveries will be reported in different parts of northeast Africa, but it might also be that in the very recent period during which the satellite tracking tech- nique has been employed, White Storks have changed their migratory behaviour.

Examination of all the available White Stork ring recoveries, subdivided into time segments, shows that during their 100-year ringing history they have clearly undergone a number of significant modifications of migratory behaviour (mainly shortening of the migra- tion distance, directional shifts in the region of the European migration divide, and an earlier return to the breeding grounds). However, there is no discernible evidence of change in their movements in the Sudan and Chad. Furthermore, specialists who have investi- gated the White Storks in the Sudan know of no such changes (C. Nikolaus & H. Schulz, pers. comm.).

On the other hand, it has long been suspected that distortions can arise from the differential reporting of ring recoveries. Schulz (1988) had this to say on the subject regarding the Sudan: ‘The distribution of ring- recovery reports and observations . . . gives the impression that the main migration route follows the Nile Valley; but the reason is that the Nile Valley is rel- atively densely populated, so that it is much more likely that storks passing through or dying in this region will be observed or found than in the unoccupied, remote regions of the Nubian Desert’. What holds for the Nubian Desert may equally apply to the relatively

sparsely populated and comparatively isolated western Sudan and Chad. Such an idea becomes especially plausible when one considers that the less sophisticat- ed inhabitants of the Sudan and other countries ‘still regard bird rings as provided by God’ and therefore tend to preserve them in shrines or wear them as amulets rather than send them back (Schulz 1988).

Following our studies, the planning of future expedi- tions and conservation projects affecting the White Stork should give priority to the western Sudan and Chad as important rest areas and intermediate destina- tions of individuals of the eastern population. In future such studies should examine the western regions with respect to the availability of food for these storks and the degree of hunting pressure on them. Only after thoroughgoing studies in the western Sudan and Chad will it be possible to assess whether the present esti- mates of losses of storks from the eastern population due to human persecution, which are of the order of 3-4% per year (Schulz 1988, Berthold 19933, are real- istic. Detailed information about the situation in the western Sudan and Chad is also extremely important with respect to a White Stork conservation project cur- rently being in progress.

After the conclusion of a six-year research pro- gramme on the migration and resting beha\.iour of storks along the eastern route, financed by the German Federal Ministry of Environmental Affairs (Querner & Berthold 1998), it is planned to ensure the preserva- tion of the White Stork populations, particularly along the migration routes as well as in the stopover areas and winter quarters, by agreement with treaty coun- tries (Minutes of a specialists’ meeting at the Ministry dated November 1999), under the auspices of the ‘Convention on the Conservation of Migratory Species of Wild Animals’ (CMS, ‘Bonn Convention’) and the UNEP. This agreement, which corresponds approxi- mately to the recently concluded ‘African-Eurasian Waterbird Agreement’ (AEWA; Interim Secretariat of the AEWA 1995), ought to specify in particular the western Sudan and Chad.

For the optimal implementation of these protective measures, studies of the stopover-site ecology of storks in the western Sudan and Chad are urgently required. These should include on-site investigation of the caus- es of the difference in flight directions of storks coming from Egypt, such that some migrate into the eastern part of the Sudan and some into the western part. The responsible factors might be endogenous, such as pro- grammed directional preferences, but they might also be external circumstances such as small-scale differ- ences in food supply.

0 2001 British Ornithologists’ Union, Ibis, 143, 450-455

Staging and wintering of White Stork 455

Finally, it has been proposed that in the Sahel zone White Storks migrating along the western route occa- sionally move into the region of the eastern migrants, or vice versa, and then later migrate back along with the storks from the other population (loop or circle migration; e.g. Nikolaus 1989). The stork we tracked from the eastern route into Nigeria provides the first evidence that eastern storks can indeed sometimes travel into the winter quarters of the western storks, which makes the above idea entirely plausible.

The project was supported by the German Federal Ministry for Environmental Affairs. For their untiring assistance in catching storks and attaching transmitters we thank M. MullerCt), Hohenauen, and Z. Jakubiez, Wroclaw.

REFERENCES

Berthold, P. 1993. Bird Migration. New York: Oxford University Press.

Berthold, P., Nowak, E. & Querner, U. 1997. Eine neue Dimension der Vogelforschung: Die Satelliten-Telemetrie. Falke 44:

Interim Secretariat of the AEWA 1995. The African-Eurasian Waterbird Agreement. The Final Act of the Negotiation Meeting

134-1 40.

(June 7995) and Agreement Text with Action Plan. The Hague: AEWA.

Jespersen, P. & Thing, A.V. (eds) 1950. Studies in Bird Migration. Being the Collected Papers of H.Chr.C. Mortensen. Munks- gaard-Copenhagen: Dansk Ornithologisk Forening.

Kaatz, M. 1995. Bearbeitung, Auswertung und Dokumentation der uber das Migrationsverhalten des WeiBstorches gewonnenen Flugzeug- und Satellifendaten. Diplomarbeit Univ. Rostock.

Nikolaus, G. 1989. Die Situation des WeiOstorchs im Sudan. In Rheinwald, G., Ogden, J. & Schulz, H. (eds) WeiBstorch - White Stork. - Proc. I Int. Stork Conserv. Symp. Schriftenr. DDA 10:

Querner, U. & Berthold, P. 1998. Satellite-tracking for the conser- vation of migratory bird species focusing on the White Stork (Ciconia ciconia). Torgos 28: 91-96.

SchOz, E., Berthold, P., Gwinner, E. & Oelke, H. 1971. GrundriB der Vogelzugskunde. Berlin-Hamburg: Parey.

Schuz, E. & Bohringer, R. 1950. Vom Zug des WeiO-Storchs in Afri- ka und Asien nach den Ringfunden bis 1949. Vogelwarte 15:

Schulz, H. 1988. WeiBstorchzug - Okologie, Gefahrdung und Schutz des WeiBstorchs in Afrika und Nahost. Weikersheim: Margraf.

337-340.

160-1 87.

Received 25 February 2000; revision accepted 12 September 2000

@ 2001 British Ornithologists' Union, Ibis, 143, 450-455