S49.5: Satellite tracking of White Storks Ciconia ciconia

¹Department of Biology, University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium, fax 32 15 413269, e-mail wbossche@uia.ua.ac.be; ²Storchenhof, Chausseestr. 18, 39279 Loburg, Germany, e-mail Kaatz@mluagis1.Landw.uni-halle.de; ³Max-Planck-Institut, Vogelwarte Radolfzell, Schloss Möggingen, Schlossallee 2, D-78315 Radolfzell, Germany.

Tagged storks remained 5 to 68 days in pre-wintering sites in Sudan and Chad. In January, White Storks were tracked in Sudan (4), Chad (2), Uganda (1), Kenya (1), Tanzania (2) and southern Africa (4). The satellite tracking program has extended our knowledge on the migration pattern and stopover behaviour of the White Stork in a spectacular way.

For long distance migrants like storks; migration pattern, stopover ecology and choice of wintering site are important features that determine survival and breeding success and are subject to selection pressure (Alerstam 1981, Bairlein 1981). We expect that the migration pattern that evolved in a species is a result of natural selection during migratory flights, when energy reserves are consumed, and during stopover periods when new energy reserves can be built up or when energy should be saved if migration conditions are not good. The migration pattern chosen by a species or individual must permit the bird to arrive at its destination safely and at a time and in a condition in which it is possible to carry out its intended activity. Storks arriving late in the breeding area have no chance to breed successfully and arrival time is dependent on latitude, migration route followed and weather conditions in the wintering area and along the migration route (Cramp & Simmons 1977, Dallinga & Schoenmakers 1989, Goriup & Schulz 1991).

White Storks are long-distance migrants using long-term soaring flight to cover large distances with low expenditures of metabolic energy. To locate lift they seem to rely on their social behaviour (Pennycuick 1972).

The popularity of storks in Europe does not protect them from the negative effects of habitat loss and habitat impoverishment (Schulz 1988). If we want to manage the stopover sites of migrating White Storks, then we have to know details about their habitat use and the factors influencing their migration pattern and wintering behaviour. In this way we will be able to determine the key-elements for their protection. This knowledge makes it possible to manage the habitats not only in favour of the storks, but also economically, with respect to both the needs of the birds and the needs of people living in the area.

Most information on White Stork movements has been gathered by ringing and field observations. But many questions about their migration pattern, stopover and wintering behaviour remained unanswered. In the 1990s it became possible to follow the movements of migrating birds by means of satellite telemetry (Nowak & Berthold 1991). The development of transmitters with a weight of 35 to 100 g made it possible to track White Storks, which have a body-mass of 3 to 4.5 kg.

White Storks were captured in Germany, Poland and Israel. The 40 German birds were equipped with transmitters in June and July at Loburg (Sachsen-Anhalt) and near Strohdene at the Gülper See (70 km west of Berlin). In northern Poland, 4 adult White Storks were captured in July 1996. In northern Israel, 9 birds were captured in August and September in the Bet She'an Valley.

From the 53 White Storks, 25 were adult and 28 1st-winter birds.

The Polish, Israeli and adult German White Storks were freeliving birds. About 80% of the 1st-winter storks were from the 'Storchenhof' Loburg. These young storks came from nests of which the parents were accidentally killed. Some young had fallen out of the nest before they could fly. They were kept in captivity for some weeks and released at the end of August and beginning of September.

The transmitters were PTT-100 from Microwave Telemetry of 30, 60 and 90 gram. The transmitter was attached to the back of the stork. On 2 storks this was done by a continuous strip of Teflon-tape and on the other birds by 3.5 mm nylon tape. The transmitters with solar-panel were placed as a backpack, as high as possible on the stork’s back, to avoid feathers covering the panel. From 1991 to 1994, all transmitters had a battery, only one was solar powered. In 1995 and 1996, most transmitters got their energy from a solar panel. The transmitters were located with the ARGOS satellite system.

The night location was measured by calculating the mean of locations between 1800 h and 0400 h GMT. The daily migration distance is the distance between night locations.

The route was divided into three regions: Europe, from Germany or Poland to the Bosphorus (41°00'N 28°30'E) in Turkey ; Middle East, from the Bosphorus to El Thor (ca. 28°00'N 33°12'E) in Sinai, Egypt ; Africa, from El Thor to the pre-wintering area in Sudan. To calculate the daily distance only data from individuals that were tracked throughout the whole region were used. From 'Prins', a male White Stork from Germany that was tracked during two autumns, only the data for 1996 were used.

To calculate the mean location of stopover sites we did not use data from individuals that stayed more than 100 km apart. We calculated the mean location of the roost site and the distance of every stork from this point. We used the proportion of storks roosting within a distance of 30 and 50 km from the mean location. The difference in the number of tagged individuals (6 to 16) in a certain area was due to different choice of stopover site and program time of the transmitter (each day or every 2 days).

We define the pre-wintering area as the location where storks finished their autumn migration. We considered the autumn migration to have ended, when storks had crossed latitude 19°N and moved less than 50 km a day during at least 5 days. Daily distance was measured between night locations. Storks were considered to be leaving the pre-wintering site when they moved more than 200 km in 2 days from the first night location.

During the course of the study we used more adult storks because of the high mortality of young storks. Only 28% of the 1st-winter satellite tracked birds reached northern Sudan. The highest mortality occurred in Europe. Most adults, 86%, reached the pre-wintering areas in Sudan.

We followed the migration route of 44 tagged storks breeding in Germany and Poland. From 40 captured in Germany, 6 followed the western route towards Gibraltar. The other 34 and all Polish storks followed the eastern route through Turkey and Egypt.

The German birds flew southeast from the region of Berlin, through the Czech Republic, central Slovakia or E Austria, W Hungary, W Romania and through Bulgaria towards the Bosphorus region in Turkey. Some birds flew a more eastern route through SW Poland, Slovakia and central Romania.

The Polish birds flew from Masuria, south through E Poland, W Ukraine, E Romania and along the west coast of the Black Sea through Bulgaria on their way to the Bosphorus region in Turkey. All 4 Polish birds crossed the Sea of Marmara near the Bosphorus.

From 22 tagged storks, 91% crossed the Sea of Marmara in the eastern sector in the Bosphorus region. Two 1st-winter storks flew across the Dardanelles in the western sector of the Sea of Marmara.

The storks flew southeast through Turkey towards the Gulf of Iskenderum, through W Syria and the Jordan Valley. From 14 satellite tracked birds, 11 entered Israel north from the Dead Sea and 3 south from the Dead Sea (Arava Valley). All 3 storks that entered Israel in the Arava Valley were tagged in 1996. From the 11 storks that entered Israel north of the Dead Sea, 5 passed through the Bet She'an Valley and continued along the Jordan Valley and 6 flew through Jordan, east of the Bet She'an Valley. They continued to Sinai (Egypt) where most birds crossed the Red Sea at El Thor. Only a few birds passed Sharm El Sheikh, which is more to the south. They reached the Nile near Qena and continued south to Sudan.

As an example we can look at the migration pattern of no 94555 an adult male from northern Germany (Fig. 1). It started its migration on 26 August 1994 and reached northern Bulgaria on night 7. During its 10th migration day it crossed the Bosphorus and needed 3 days to fly through Turkey. This male flew fast across Syria, Jordan and Israel and slept on night 15 already on the Egyptian mainland. It reached northern Sudan on day 18.

Tagged storks returned in spring along the eastern migration route and followed in general the same route in Africa and the Middle East as in autumn.

In 1994, no 94555 was tracked from Chad to Suez (Egypt) and 'Prinsessin' (no 94552) from Tanzania to Germany. In 1996, 3 Polish storks started from Kenya (no 96551), Botswana (no 96552) and Sudan (no 96992) and were tracked on their spring migration.

In Sudan, all 4 tagged storks arriving from the south avoided the Dongola bend in the Nile and crossed the desert towards the east side of Lake Nasser.

'Prinsessin' crossed the Sea of Marmara at the Dardanelles instead of at the Bosphorus, as it did in autumn. To reach its breeding area near Berlin, it flew through Ukraine and Poland which was more to the east than in autumn.

19 tagged White Storks followed in 1992-96 reached their pre-wintering grounds in Sudan and Chad.

The number of days they needed to reach Sudan is shown on Figure 2. The first day on which the birds flew 50 km or more in the general migration direction was taken as the starting date. During stopover days, the distance between the two night locations was less than 50 km. Storks did not migrate during the night.

As Figure 2 shows, 4 1st-winter and 9 adult storks covered the distance from the breeding area to northern Sudan rapidly in a mean period of 20 days, stopover days included. There were 3 individuals that made long stopovers in the Middle East. One adult took 31 days and a 1st-winter 66 days from Germany to Sudan. One adult flew from Poland to Sudan in 57 days.

Some storks stopped already for more than one night after only one or two migration days in Europe. Two storks stopped for a longer period of 5 and 22 days near the Bosphorus. Most migrants flew fast through the Middle East. One stork stayed for a longer period in Turkey, one in Israel and one in Saudi Arabia.

While most storks stopped only for a short period and mostly only once, some individuals (no 95550, 95554, 96992) stopped several times, even for longer periods. This could be a reflection of the poor condition these birds were in. We have no data on the condition of these birds prior to their departure.

We could follow 4 tagged storks during at least part of their spring and autumn migratory flight. 2 could be followed along their whole migration route from and to the breeding area.

These 2 storks needed only 21 days to fly from the breeding area to Sudan, but took more than twice as much time to cover the same distance in spring. No 96551 flew in 51 days from Sudan to Poland and no 94552 flew in 49 days from Sudan to Germany. They made more and longer stopovers in spring and covered shorter daily distances on spring days with bad weather conditions.

In Europe, 70% of young storks and 57% of the adults stopped at least once for more than one night at a location. Because many of these stops occurred soon after departure this might be influenced by bad weather conditions along the route. In central Europe, tagged storks were in small flocks and were attracted to tractors working on fields to feed in their vicinity. In the Middle East, only 20% of young and 36% of adult storks stopped for more than one night.

If a certain location is important as a stopover site, we expect many storks to try to land in different years at this place. Thus when we look at tagged storks we expect the mean distance between night locations of individuals to be small and the proportion of birds in a short radius (<30 km) to be high.

The mean location of the one night stopover sites is based on data from satellite tracked birds (Fig. 3). We plotted the night location of 27 migrating storks and those grouped at a certain longitude and latitude were used for the analysis.

At the bottleneck areas at the Bosphorus, Goksu delta and Iskenderum tagged storks roosted at traditional sites with 50 to 76% of the individuals roosting within a radius of 50 km. At other areas, the distances between night locations were larger.

'Prins' roosted in Europe at different places in both autumns. After crossing the Bosphorus it slept 41 NW of the site in 1994, but in the Middle East it stayed 86 to 125 km from previously used sites. In Sudan the distance between the sites was on two occasions quite small, 34 and 55 km.

As an example we can look at the migration of the male no 94549 which migrated from Germany on 1 September 1994. It stopped 55 km E from its starting point for one day (Fig. 4). On 3 September, it flew 342 km SE and continued towards the Bosphorus covering even 504 km on 6 September. It slowed down when it reached the Bosphorus and crossed it on 8 September. It arrived at Iskenderum on 11 September and flew through Syria to arrive on 14 September in the Bet She'an Valley in Israel. It stayed in the valley for 12 nights and continued migrating to the south, crossing the Red Sea near El Thor on 27 September. When it reached the northern point of Lake Nasser it flew only 70 km on 29 September, but continued its migration to central Sudan.

Inside Europe the tagged storks covered a mean daily distance of 218 km for adult and 1st-winter birds (Table 1). This is short compared to the mean daily distance of 275 km and 288 km in the Middle East and Africa.

In a model which tests for year, region and age; only region significantly affected variation in daily migration distance (Repeated ANOVA: F_2,37.8 = 13.43, P < 0.001).

The range of daily migration distance in autumn for all individuals is listed in Table 2. Data from birds that could not be tracked through the whole region were also used. Data from the winter period are not included.

White Storks migrated up to 500 km per day, in Europe as well as in the Middle East and Africa.

During spring migration, tagged storks migrated up to 550 km per day in Africa and in Europe and Middle East more than 350 km per day (Table 3).

During 1992-96, we collected information on the pre-wintering areas of 18 tagged White Storks. Only 4 of them were 1st-winter birds.

Eastern pre-wintering areas were located east from the White Nile at Wad Medani and northeast from the Blue Nile in the Shukriya region, both in Sudan (Fig. 5). Central pre-wintering areas were located west of the White Nile in a large area around the Nuba mountains in Sudan. Western pre-wintering areas were located in Ouaddai, E Chad and in Darfur, W Sudan.

Tagged White Storks stayed between 5 to 68 days at a pre-wintering site. From 3 storks the transmitter failed. The number of days spent at the pre-wintering site varied between years (ANOVA: F_4,8.27 = 6.59, P = 0.01). Region, age and arrival day had no significant influence (all P>0.05).

The tagged male 'Prins' (no 94550-96991) chose in 1996 a pre-wintering site 213 km southeast of its 1994 site.

The location of the pre-wintering areas in the east, west or central zone was partly visible in the stork’s migration direction, when they passed along the east side of Lake Nasser at car 23°00'N 33°00'E (Point A on Fig. 5).

8 tagged storks continued due south from the east side of Lake Nasser crossing the Nubian Desert towards 20°00'N 33°00E (from A to B on Fig. 5). Of these 8 birds, 3 changed their direction to west-southwest when they again approached the Nile at car 20°00'N 32°50'E and flew to a western pre-wintering area. The 5 other storks flew to a pre-wintering area in the east.

6 tagged storks continued their migration along or close to the western corner of Lake Nasser further southwest and stayed close to the Nile (from A to C on Fig. 5). From these 6 storks, 3 continued to a central pre-wintering area and 3 to a western site. None of these birds flew northeast to follow the Nile upstream at Merowe (point C on Fig. 5, ca. 18°20'N 31.45'E).

After leaving the pre-wintering area, tagged storks stayed in Sudan (4) or flew to Chad (2), Uganda (1), Kenya (1), Tanzania (2) and southern Africa (4). From the storks wintering above the equator, 3 in Sudan, one in Uganda and one in Chad were not tracked during the complete wintering period.

The migration routes followed by tagged storks correspond with previous observations of migrating storks. Most tagged individuals followed narrow routes through the Middle East up to Lake Nasser in southern Egypt.

White Storks avoid crossing large water bodies and pass along narrow migration flyways in Spain and the Middle East. At these bottleneck areas, storks still have to cross several kilometres of water. In Sinai, the Gulf of Suez is 25 to 30 km wide, in Turkey the Bosphorus is only 1 to 5 km wide and the Gulf of Iskenderum up to 30 km. In Spain they cross the Strait of Gibraltar, which is 14 to 20 km wide. White Storks cross water bodies by gliding from high altitudes and flapping low above the water (Koch et al. 1966, Heckenroth 1968). Migration of White Storks across the Mediterranean Sea is rarely observed (Bijlsma 1987).

Based on ringing recoveries and observations, about 25% of the storks cross the Sea of Marmara (Kasparek & Kilic 1989). Only 2 out of 22 tagged storks chose this route in autumn.

Larger numbers use the narrow sea crossing at the Bosphorus in autumn (Bijlsma 1987). From the Bosphorus, most storks migrate southeast and reach the Mediterranean at the Göksu Delta. From there they follow the shore line (Kasparek & Kilic 1989). At the northeastern point of the Mediterranean, storks turn south at the Gulf of Iskenderum and follow as guidelines the Orontes, Litani and Jordan Valley towards the southern point of the Dead Sea (Schulz 1988). Few pass through Lebanon (Kumerloeve 1961). The majority migrate through the Jordan Valley to the Dead Sea and cross the central Negev to northern Sinai. Almost no storks enter Africa through Suez (Schüz 1955), but cross the southern part of the Gulf of Suez near El Thor (Koch et al. 1966, Safriel 1968). All tagged storks crossed the Red Sea in southern Sinai. Small numbers cross the Gulf of Suez at Ras Mohammed, the southern point of Sinai. At this point there are islands that could attract the migrating flocks (Koch et al. 1966).

In the Middle East, small number migrate east of the Rift Valley (Heckenroth & Schüz 1970).

In Africa, White Storks continue their migration west of the Rift Valley, through the Lake Victoria basin. This route is also followed by the Abdim's Stork Ciconia abdimii (Pomeroy 1989).

The spring migration route is from northern Sudan the same as the autumn route. In Sudan, storks might follow the Nile near Khartoum, but cross near Abu Hamed the desert towards Korosko, avoiding the Dongola bend (Schüz 1959, Reed & Lovejoy 1969). In Egypt, most storks follow the Nile from Abu Simbel (now Lake Nasser) until Qena (Reed & Lovejoy 1969). There most fly northeast to the Red Sea and cross near El Thor (Schüz 1955, Reed & Lovejoy 1969). Small numbers passing at the end of spring migration keep following the Nile up to the delta and cross near Suez. These are mainly 1st-winter, immatures and weak adults (Schüz 1955). Schüz (1955) calls the Nile Delta the 'stork hospital'.

Migration close to the Mediterranean coast through the western Negev and west of Be'er Sheva only occurs in spring (Schüz 1955). The main migration route goes from Aquaba to the Dead Sea (Schüz 1955, Schulz 1988).

The route is fixed because of the location of the Mediterranean and orientation of valleys (Koch et al. 1966). Besides these geographical guidelines, the Nile attracts migrating birds. In Egypt and northern Sudan there is only a small zone with vegetation close to the river, which the storks follow (Mathiasson 1963).

Data from tagged storks confirm ground observations. In normal autumns, about 50% of all storks migrating through Israel can actually be seen in the Bet She'an Valley. This was also observed during two synchronized surveys along the eastern migration axis from 2 to 7 September 1990. One survey covered the Bet She'an Valley and a second was held more to the south in part of the eastern Negev and Arava Valley, near Dimona. During this period, 143,572 White Storks were counted in the south and only 96,112 in the north (Leshem 1991).

Satellite telemetry has greatly extended our knowledge of the migration pattern of White Storks. The majority of tagged storks tried to reach pre-wintering areas in Chad and Sudan as fast as possible.

Schüz (1952) stated that storks fly slower in autumn, covering a daily distance of 100 km. He observed that they left the eastern breeding grounds in the 3rd decade of August and only arrived at the end of November in southern Africa. However in later publications he wrote that White Storks migrate fast because there is no food available in the Middle East (Schüz 1955).

Some long-distance migrating species have to stop several times along the migration route in order to complete the journey successfully. According to Moreau (1972), White Storks migrate slowly in autumn, allowing them to feed at various locations along the route. In spring, they migrate relatively quickly (Moreau 1972). This does not correspond with the observations from tagged storks.

In spring, tagged storks flew fast through areas with good thermal conditions in Africa and Sinai. There was one exception: male no 94555 flew very slowly and followed the Nile until Suez. This route is mainly followed by young or weak storks (Schüz 1955).

In spring storks migrate fast, without long stops (Schüz 1952 and 1955, Schulz 1986). It is only the non-breeding and sick storks that stay long, breeding adults fly rapidly to the breeding area (Schüz 1955). In March the feeding conditions are normally good after the rain in desert areas such as the Negev and in mountain regions such as the Golan Heights. These good feeding conditions attract storks to land where they feed for long periods and fly only short distances (Hutson in Schüz 1955). Some flocks might even rest for some days in Israel (Schulz 1988).

Weather conditions can differ widely between years and at various locations along the route and might explain these contradictary observations.

Because White Storks migrate fast, without long stops, they must start their autumn migration with sufficient energy reserves. Their soaring behaviour on migration reduces energy use (Pennycuick 1972). Their long migration days of 8 to 10 hours (Michev & Profirov 1989, Leshem & Yom-Tov 1996), leaves them little time to look for feeding sites. In addition, the large size of autumn flocks in the Middle East, up to 20,000 birds, makes it difficult for an individual to find enough food at stopover sites.

Along most parts of their migration route from Europe to Africa, tagged storks roosted at different sites. Only near bottleneck-areas at the Bosphorus and Iskenderum many birds were roosting at 'traditional' sites.

White Storks by-pass possible stopover sites in autumn, while in spring more birds will stop and try to feed. White Storks are time-minimisers, but are also energy-minimisers because they rely almost exclusively on thermals.

Tagged storks flew on autumn migration mean daily distances of 218, 275 and 288 km in Europe, the Middle East and Africa. At maximum they flew 504 km on one day. We observed no differences between adults and 1st-winter individuals.

This difference can be caused by less favourable weather conditions in Europe. Some birds rested one or more days soon after they started their migration. Rain and wind can cause bad thermal conditions, so that the storks have to interrupt their migration. Until they reach Bulgaria, most flocks are small, containing tens to exceptionally hundreds of birds (M.Kaatz, pers.com.). This could allow the birds to make more and longer feeding or resting stops during the day, which shortens the daily flying distance. The tagged male 'Caesar' (no 94555) was followed by glider and flew in Europe on 27 August 1994 and 29 August 1994 for a shorter period of about 4.5 h, covering only 115 km and 88 km. On 28 August 1994, it flew for 9.40 h and covered a distance of 334 km (Kaatz et al. 1996).

Our observations are comparable with data collected with radar and motor-gliders. Michev and Profirov (1989) calculated a daily autumn migration distance of 317 km over land, based on a migration speed of 38.2 km/h and a mean migration day of 8.18 h of groups tracked by radar near Burgas in Bulgaria. The 14 satellite tracked White Storks that passed near Burgas (42°30'N 27°25'E) in autumn covered a daily distance of 341 km (S.D.= 91).

Leshem and Yom-Tov (1996) followed flocks with a glider, mainly during spring migration, and found a mean velocity of 38.7 km/h. With a mean soaring time of 9 h, the storks covered a mean distance of 348 km. The 14 satellite tracked birds that passed near the Dead Sea (31°00'N 25°20'E) in Israel in autumn, covered a daily distance of 298 km (S.D.=81).

Most storks are ringed as nestlings. This means that the ringing date is not the date a stork started its autumn migration. Based on ringing recoveries, autumn migration speed was calculated between 100 and 200 km (Haverschmidt 1949). One German stork left its nest on 3 August 1933 and was recovered 2 nights later in Italy, about 610 km distant (Haverschmidt 1949). Stork 'Bob' with ring Helgoland 2188, at that moment a 7th-winter bird, left Zimbabwe (18°16'S 29°55'E) at 1 February 1979 and arrived in Germany (53°39'N 7°43'E) on 28 March 1979 (Meybohm & Fiedler 1983). This stork flew about 9012 km in 55 days, with a mean daily distance of 164 km.

Satellite telemetry has provided for the first time reliable information on the daily migration speed of storks of different ages at various locations.

The start of the migration is an endogenously stimulated behaviour during the regular autumn migration period (Berthold 1996). The place where 1st-winter birds will end their autumn migration can be determined by several factors. This can be socially induced when adults might take young birds to a wintering site (Berthold 1996) or young Starlings Sturnus vulgaris might stop where food is abundant (Perdeck 1964). An individual migrating bird might stop after flying a certain distance or time, the length and time might differ for species and populations (Berthold 1996).

While the regular wintering areas of the White Stork are found south of 14°N (Schüz 1959, Schulz 1988), all tagged storks stopped for several weeks in Sudan or E Chad between 16 to 10°N and 19 to 36°E.

Based on observations and ringing recoveries, Schulz (1986) located the main pre-wintering area in Sudan, south of Khartoum, between 16° and 12°N and 28° to 36°E. Most observations and most autumn recoveries were in the grass savanna east of the White Nile, especially near Wad Medani, towards the border with Ethiopia. These areas are used yearly and the borders are quite fixed, but numbers differ between years (Schulz 1988). West of the White Nile, there are less recoveries and fewer observations. Only near the area of Jebel Marra were storks regularly observed. In the region of Nyala and Bahr-el-Arab and Bahr-el-Ghazal even a few thousands could winter (Hogg et al. 1984). During a survey in September-October 1986, he observed 3,000 storks along the road in Wad Medani and 6,000 between the White and Blue Nile.

In W Sudan, in the provinces of Kordofan and Darfur, most storks were observed passing through without resting for a longer time (Schulz 1988).

The results from satellite telemetry showed that W Sudan and E Chad are more important pre-wintering areas than predicted. From 17 tagged storks, only 5 stayed east of the White Nile.

The choice of pre-wintering area is dependent on food availability -mainly grasshoppers-, low vegetation -savanna, farmland- and presence of water -pools or river- (Schulz 1986). South of 10°N in the floodplain of the sudd, the vegetation could be to high to feed in.

Tagged White Storks arrived in the pre-wintering area in the middle of September and left the area from the end of October till to middle of December. Field observations and ringing data show that they arrive in central Sudan from August until early October (Schüz & Böhringer 1950, Schulz 1986). From the middle of October, storks start leaving Sudan to the south, although some individuals stay all winter (Schulz 1988). In 1986, the departure coincided with a change in wind direction from S to N (Schulz 1988). From November to February, ringed 1st-winter birds were reported from Uganda to South Africa (Schüz & Bohringer 1950). These observations correspond with the arrival of storks in Kenya at the end of October and beginning of November (Pomeroy 1989, Schulz 1988).

The number of days tagged storks roosted varied between years and region; age or arrival date had no influence. Storks feeding in southern Sudan and Chad are dependent on locusts, of which the numbers are influenced by the amount of yearly rainfall (Dallinga & Schoenmakers 1989). Schulz (1986) observed how flocks changed feeding and roosting areas by more than 10 km in a few days. In an area with densities of 30 grasshoppers/m², there were a few days later only 0.5 grasshoppers/m². The storks chose the sites in relation to the presence of food.

The distribution of eastern White Storks in Africa is at least in some winters related to outbreaks of pest species.

In January 1987 in an area of 25 km² in Tanzania 100,000 storks were feeding on massive numbers of Army worms. Almost no storks were seen in southern Africa that winter (Schulz 1988). In years with extensive rains and thus higher food availability in E Africa, White Storks arrive later and in smaller numbers in southern Africa (Dean 1964). In years with dry conditions in northern parts of E Africa, up to 200,000 could 'winter' in South Africa (Schulz 1988). For example, large numbers arrived in 1972/73 in Natal, South Africa (Maclean et al. 1973). In spring 1973, White Storks arrived late and in smaller numbers in central Europe (Schüz & Szijj 1975).

In southern Africa, the regional distribution of recoveries is concentrated in the summer-rainfall area (Oatley & Rammesmayer 1988). The expansion of storks into the SW Cape coincided with the introduction of alfalfa as a ley crop.

Facultative movements in winter seem to be exogenously controlled by food availability (Berthold 1996). White Storks left the pre-wintering area at different times and between years there are large differences in the proportion wintering in central or southern Africa (Dallinga & Schoenmakers 1989). Taking into account their gregarious behaviour and concentrations near grasshopper outbreaks it is unlikely that many storks show a high fidelity to their winter feeding sites in Africa.

Wintering in central or southern Africa results in a delay in spring of at least 14 days in the latter case. In addition the birds might arrive in a poorer condition because of the longer distance flown.

There are contradictary data on the existence of age-related wintering areas of eastern storks. In Uganda and Kenya, Schüz and Bohringer (1950) found a proportion of 56% 1st winter storks, recovered in December and January. In southern Africa however, 71% of recovered storks were in their 1st-winter. According to Cramp and Simmons (1977) and Oatley & Rammesmayer (1988) , the proportion of 1st-winter storks is higher in South Africa than further north. In contrast, Kania (1985) found no differences between wintering grounds of Polish storks of various ages recovered in December-January. Schmidt (1987) also found no age differences in wintering areas of German storks.

Many aspects of White Stork migration and location of wintering grounds remain unclear. Further research with tagged storks should concentrate on summering grounds of 1st-summer storks and the relation of these places with their future breeding site. Further data are needed on the repetitive choice of wintering areas of storks of different age and geographical origin.

The White Stork project has been supported by the Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit, Bonn and the Society for the Protection of Nature in Israel. Thanks to Peter Berthold for commenting on an earlier draft.

Alerstam, T. 1981. The course and timing of bird migration. In: Aidley, D.J. (eds) Animal migration. Cambridge University Press, Cambridge.

Bairlein, F. 1981. Analyse der Ringfunde von Weisstörchen Ciconia ciconia aus Mitteleuropa westlich der Zugscheide: Zug, Winterquartier, Sommerverbreitung vor der Brutreife. Die Vogelwarte 31: 33-44.

Berthold, P. 1996. Control of Bird Migration. Chapman & Hall, London.

Bijlsma, R. 1987. Bottleneck areas for migratory birds in the Mediterranean region. ICBP Study report 18, 135 pp.

Cramp, S. & Simmons, K.E.L. (eds) 1977. The birds of the Western Palearctic, Vol 1. Oxford University Press, Oxford.

Dallinga, H. & Schoenmakers, M. 1989. Population changes of the White Stork Ciconia ciconia since the 1850s in relation to food resources. In: Rheinwald, G., Ogden, J. & Schulz, H. (eds) Proceedings 1st International Stork Conservation Symposium. Schriftenreihe des Dachverbandes Deutscher Avifaunisten 10. Braunschweig: 231-262.

Dean, G.J.W. 1964. Stork and Egret as predators of the Red Locust in the Rukwa Valley outbreak area. The Ostrich 35: 95-100.

Goriup, P.D. & Schulz, H. 1992. Conservation management of the White Stork. In: Salathé, T. (ed.) Conserving Migratory Birds. ICBP techn. publ. 12, Cambridge: 393 pp.

Haverschmidt, F. 1949. The Life of the White Stork. E.J. Brill, Leiden, The Netherlands.

Heckenroth, H. 1968. Beobachtungen über die Rechtswendung des wegziehenden Weisstorchs am Golf von Iskenderun. Die Vogelwarte 24: 246-262.

Heckenroth, H. & Schüz, E. 1970. Funde in Europa beringter Weisstörche im Orient östlich der Schmallfront. Zool. Abh. Staat. Mus. für Tierk. in Dresden 31: 193-203.

Hogg, P., Dare, P.J. & Rintoul, J.V. 1984. Palaearctic migrants in the central Sudan. Ibis 126: 307-331.

Kaatz, M., Dubian, K., Rehbein, R. & Schaffer, T. 1996. Zugbegleitung besenderter Weisstörche - Neue Schutzmöglichkeiten auf den Zugwegen. In: Kaatz, Ch. & M. (eds) Jubiläumsband Weisstorch. Tagungsbandreihe des Storchenhofes Loburg im MRLU-LSA, 3. Tagungsband: 129-133.

Kania, W. 1985. Results of bird ringing in Poland. Migrations of White Stork Ciconia ciconia. Acta Orn.: 1-41.

Kasparek, M. & Kilic, A. 1989. Zum Zug des Weisstorchs durch die Turkei. In: Rheinwald, G., Ogden, J. & Schulz, H. (eds) Proceedings 1st International Stork Conservation Symposium. Schriftenreihe des Dachverbandes Deutscher Avifaunisten 10. Braunschweig: 297-306.

Koch, A., Magnus, D., Seilkopf, H. & Baron, H. 1966. Der Weisstorch-Zug im Raum Sinai bis Kena in landschafts-morphologischer Sicht. Die Vogelwarte 23: 209-220.

Kumerloeve, H. 1961. La Cicogne blanche en Anatolie, Syrie, Liban et Palestine. Alauda 29: 25-29.

Leshem, Y. 1991. Study and forecast of fall and spring migration of soaring birds over Israel and the effect of climate factors on migration. Thesis submitted for the degree 'Doctor of Philosophy'. Tel Aviv University, November 1991.

Leshem,Y. & Yom-Tov, Y. 1996. The use of thermals by soaring migrants. Ibis 138: 667-674.

Maclean, G.L., Gous, R.M. & Bosman, T. 1973. Effects of Droughts on the White Stork in Natal, South Africa. Die Vogelwarte 27: 134-136.

Mathiasson, S. 1963. Visible Diurnal Migration in the Sudan. Proc. XIII Int. Orn. Congr. 1963: 430-435.

Meybohm, E. & Fiedler, G. 1983. Neue Fälle von hohem Alter, Ortstreue, Um- und Fernsiedlung und anderen brutbiologischen Befunden beim Weisstorch Ciconia ciconia. Die Vogelwarte 32: 14-22.

Michev, T. & Profirov, L. 1989. Der Herbstzug des Weisstorches entlang der Bulgarischen Schwarzmeerküste. In: Rheinwald, G., Ogden, J. & Schulz, H. (eds) Proceedings 1st International Stork Conservation Symposium. Schriftenreihe des Dachverbandes Deutscher Avifaunisten 10. Braunschweig: 281-295.

Moreau, R.E. 1972. The Palaearctic-African bird migration systems. Academic Press, London.

Nowak, E. & Berthold, P. 1991. Satellite tracking: a new method in orientation research. In: P. Berthold (ed.) Orientation in Birds. Birkhäuser, Basel: 307-321.

Oatley, T.B. & Rammesmayer, M.A.M. 1988. Review of recoveries of ringed White Storks Ciconia ciconia in southern Africa. The Ostrich 59: 97-104.

Pennycuick, C.J. 1972. Soaring behaviour and performance of some East African birds, observed from a motor-glider. Ibis 114: 178-218.

Perdeck, A.C. 1964. An experiment on the ending of autumn migration in starlings. Ardea 52: 133-139.

Pomeroy, D. 1989. The White Stork Ciconia ciconia in Kenya. In: Rheinwald, G., Ogden, J. & Schulz, H. (eds) Proceedings 1st International Stork Conservation Symposium. Schriftenreihe des Dachverbandes Deutscher Avifaunisten 10. Braunschweig: 355-356.

Reed, C.A. & Lovejoy, T.E. 1969. The migration of the White Stork in Egypt and adjacent areas. The Condor 71: 146-154.

Safriel, U. 1968. Bird migration at Eilat, Israel. Ibis 110: 283-320.

Schmidt, B. 1987. Auswertung der Ringrückmeldungen des Weisstorchs. WWF/ICBP Weisstorchprojekt, unveröff. Gutachten.

Schulz, H. 1986. Untersuchungen zur Gefährdung des Weisstorches auf den Zugrouten und in der Überwinterungsgebieten. Bericht Forschungsaufenthalt Sudan, sep/okt 1986. Umweltstiftung WWF-Deutschland, ICBP.

Schulz, H. 1988. Weisstorchzug. Ökologie, Gefährdung und Schutz des Weisstorchs in Afrika und Nahost. Weikersheim: 459 pp.

Schüz, E. 1955. Vom Zug des Weisstorchs im Raum Syrien bis Ägypten. Die Vogelwarte 18: 5-13.

Schüz, E. 1959. Problems about the White Stork Ciconia ciconia in Africa seen from a European viewpoint. Ostrich, suppl. 3 (Proc. 1st Pan African Ornithol. Congr. 1957): 333-341.

Schüz, E. & Böhringer, R. 1950. Vom Zug des Weisstorchs in Afrika und Asien nach den Ringfunden bis 1949. Die Vogelwarte 15: 160-187.

Schüz, E. & Szijj, J. 1975. Bestandsveränderungen beim Weisstorch, fünfte Übersicht: 1959-72. Die Vogelwarte 28: 61-93.

Steiof, K. 1987. Massenzug des Weisstorchs Ciconia ciconia bei Hurghada/Ägypten. Die Vogelwarte 34: 100-106.