In Europe, increasing numbers of Slender-billed Gull Larus genei have been observed away from the breeding colonies in recent years. This paper describes this extralimital occurrence, both in space and in time, by analyzing the species’ migration strategies, dispersal behaviour and vagrant records. In order to establish a picture of all records of vagrant Slender-billeds, reports of national rarities committees were consulted and, in addition, information was asked from several committees (vagrant reports were not taken into account when not accepted by the relevant committee).
Distribution and status
Slender-billed Gull breeds locally (and rather disjunct) in coastal areas of West Africa (c 22 500 individuals), the Mediterranean and Black Sea (123 000-237 000 individuals, and eastwards into Central Asia and north-western India (c 150 000 individuals) (Delany & Scott 2002). Since the 1960s, it has spread from the Black Sea through the eastern Mediterranean to the western Mediterranean (Serebryakov & Zubakin 1997). More than 90% of the total number of European breeding birds are concentrated at just a dozen sites (BirdLife International 2004, figure 1).
The current European breeding population is concentrated in Italy, Turkey and Ukraine, with smaller populations in Azerbaijan, France and Spain (BirdLife International 2004). The latest published figure for Italy is 3350 pairs in 2001 (Serra & Brichetti 2004). The Spanish breeding population is estimated at 800-900 pairs (Martí & del Moral 2003). In France, c 850 pairs breed, all in the Camargue, Bouches-du-Rhône (Sadoul et al 2003).
Migration
The European populations have developed different migration strategies. The western Mediterranean populations winter in the south of their breeding range and both the Atlantic and Mediterranean coasts of north-western Africa, partly together with the western African population (Olsen & Larsson 2004). Birds of the Black Sea population spend the winter on the northern African coast from Egypt west to Tunisia but also in Bulgaria, Greece and Israel (Olsen & Larsson 2004). Most ringing recoveries in Tunisia, for example, relate to birds from the Black Sea population. After the species extended its breeding range into the western Mediterranean, French, Italian and Spanish birds were also increasingly observed there. Individuals ringed in Tunisia have also been recovered in Egypt and Libya (Isenmann et al 2005). The species returns to its Black Sea and western Mediterranean breeding sites from late March, with adults peaking in April and early May and immatures in May (Dies & Dies 2000, Olsen & Larsson 2004).
Most movements outside the breeding season occur within the species’ breeding range, with the exception of those by the populations from West and Central Asia that winter around the Arabian Peninsula, east to western India (Glutz von Blotzheim & Bauer 1982, del Hoyo et al 1998). Birds leave their breeding grounds in July (Olsen & Larsson 2004).
Dispersal and immigration
The species mostly breeds in dynamic habitats: shallow saline areas or salt marshes where conditions may vary with or within breeding seasons. In the western Mediterranean, breeding colonies are regularly found on artificial islands in salinas. Modern management of these salinas involves levelling each salina to maximize the surface water area. Ground-nesting birds, like Slender-billed Gull, are thus regularly displaced from their breeding sites. Also, natural breeding sites are being converted for agriculture or recreation (Tucker & Evans 1997). From year to year, numbers (as well as breeding sites) tend to fluctuate accordingly (Oro 2002, Serebryakov & Zubakin 1997), causing dispersal.
Within a colony, the females (and their mates) in the best breeding condition have been found to select the most suitable breeding site (Oro 2002). This is another cause for the high breeding site turnover in this species, which is confirmed by several ringing programmes.
Individuals ringed in France have been found breeding in colonies in Italy and Spain, Italian ringed birds were found breeding in France and Spain, while birds ringed in Spain have been recorded in France. For example, in a newly established colony in Aude, southern France, in 2004, birds were controlled that had been ringed in Italy (one) and Spain (two). The Italian bird was the oldest in the colony; it had been ringed 12 years earlier in the Po delta, Veneto, Italy. Since 1999, it had been controlled as a breeding bird only in the Camargue, Bouches-du-Rhône (Gonin 2004). The Spanish population, however, shows a higher degree of philopatry. Most recoveries of birds ringed in the major colonies at the Ebro delta, Catalunya, and Coto Doñana, Andalucía, are local with only some individuals recovered from France and Italy (Manuela Forero in litt).
In Italy, c 2500 Slender-billed Gulls have been ringed (Po delta, Apulia and Sardinia) in 1991-2004, which resulted in more than 1400 resightings of 666 individuals. Most were observed in or near Italian breeding colonies, as well as in France and Spain, and during winter in Tunisia. Interestingly, one individual ringed as pullus in Foggia was observed in N’Djamena, Chad, the following winter (Nicola Baccetti in litt).
Vagrancy in Europe
In France, observations away from the Mediterranean coast have increased corresponding with the exponential increase in numbers of breeding pairs in the Camargue. Most of these records are from the valley of the Rhône and its tributaries (up to northernmost France), with some birds reaching the Atlantic coast (Dubois et al 2000) where, in 2004, the first (unsuccessful) breeding attempt took place on Île de Noirmoutier, Vendée (Desmots et al 2004). Interestingly, the majority of these vagrants in Europe were recorded at inland sites. As with many French vagrants, birds found in Central Europe may have followed rivers like the Rhône and the Po (and their tributaries) and from here to, for example, Austria, Germany and Switzerland. Other birds probably follow the (Atlantic) coast lines and end up in the North Sea area.
In Spain, however, the species is very rare on the Atlantic coast, with only a handful of records for Asturies and the Basque country (Ricard Gutiérrez in litt). In Portugal, the species is being assessed since 1995, when the Portuguese rarities committee (Comité Português de Raridades, CPR) became operative. During 1995-2003, 14 records of 71 individuals have been accepted. It is very rare on the west coast but probably a scarce migrant (and rare winter visitor) to the eastern Algarve, especially Castro Marim. It is not unlikely that in the near future the CPR will stop considering the species (Gonçalo Elias in litt). The Portuguese records have therefore not been included in table 1. On the Croatian and Slovenian side of the Adriatic Sea, the species is surprisingly rare, given the rather close proximity of Italian populations. However, recent records (see table 1) may suggest that the species is probably a very scarce visitor (Komsija za Redkosti 1993, Rubinič et al 2002, Sovinc 1993).
The French and Spanish breeding populations have increased by more than 80% over 1990-2000 (BirdLife International 2004). It is clear from figure 1 that the number of extralimital individuals in Europe has accordingly increased dramatically over the past decade. The French and Spanish populations (together with the growing Italian population) are also the likely sources for vagrancy north of the European breeding range. The fact that both of the ringed birds found in Switzerland (see table 1) had been ringed in the Camargue seems to confirm this.
The first three British records (see table 1) are very interesting in this respect, because they occurred at a time when the European breeding population was concentrated around the Black Sea, while breeding in the western Mediterranean was irregular, involving very small numbers. Likewise, records from Belarus, Finland, Poland and Sweden and some of the Hungarian records seem to confirm the notion that some European vagrants originate from eastern populations (SOF 2003). On that note, Cramp & Simmons (1983) mention a bird ringed as a chick in the Ukraine and found on the Canary Islands. However, this record is considered suspect due to the possibility of misidentification with Mediterranean Gull L melanocephalus in the breeding colony (if the chick is seen away from the parents). The species is currently considered a vagrant in the Canary Islands, with only a few records (Ricard Gutiérrez in litt). Interestingly, a bird ringed in Coto Doñana in 2001 was observed on Fuerteventura, Canary Islands, in January and March 2002. In March 2004, it was again observed and photographed there (Michiel Versluys in litt) and, the following month, it was observed at Oued Souss, Morocco (Manuela Forero in litt).
Away from France, the increase in the number of vagrants in Europe started with the influx in Switzerland in 1997. Until that year, records were widely scattered around Europe but since 1997 have been concentrated in central Europe (Austria, Germany, Hungary and Switzerland) and France (table 1). The reasons for this changing pattern are unclear. Data on French and Spanish colonies indicate that nothing out of the ordinary happened in the years leading up to 1997 (Daniel Oro in litt). Further influxes took part since then and particularly in 2001, which was most pronounced in France and Switzerland and, to a lesser degree, Austria. Also, until the invasion of 1997 mainly individual (in a few exceptions more than two) birds were recorded (see figure 1). Since that year, group size increased dramatically. Besides Portugal, the largest groups were recorded in Switzerland (20 in 1997, 11 in 2001 and 10 in 2002) and France (17 in 2001).
Of a grand total of 205 individual vagrants in Europe (78 records), only 11 first-summer and two second-summer birds have been recorded. Most observations are of birds that stay for only a very short period, usually one day, before moving on. All groups consisted of adults only. Since many immature birds stay on the wintering grounds (Malling Olsen & Larsson 2004), this is not unexpected. This could indicate that mostly birds in breeding dispersal are involved, as is evident from observations of displaying pairs in Austria, Britain and France. Of the seven birds present in Germany in May 1997 (see table 1), two pairs showed courtship behaviour and even copulated. The notion of breeding dispersal is further augmented by the seasonal pattern of extralimital European records, with the great majority of birds (151) recorded in May. A notable exception is the influx in France in 2000, which took place in April (see table 1). The total for April is 33 individuals (16 records). The few late June-July records of adults probably relate to post-breeding dispersal (or failed breeders). No pattern is evident for records of immature birds but (as far as the few data allow) their occurrence seems to mirror that of adults, although with a slight emphasis on April.
Vagrancy outside Europe
In Asia, vagrants have reached Nepal and Sri Lanka (Malling Olsen & Larsson 2004). Further east, Slender-billed Gulls have been recorded in South Korea (an adult-winter at Gwangyang Bay, Jeollanam, on 9-11 January 2002; Moores 2002) and Japan (two birds wintering in Fukuoka City, Fukuoka, with one returning to the same areas between 1984 and 1992 and the second (or another) bird in Kitakyushu, Fukuoka, in the winter of 1991/92 (Brazil 1991, Moores 2002). Three records have been accepted for Hong Kong, China; an adult on 10 and 25-26 February 1990, a first-summer on 3-10 April 1992 and an adult on 27-28 February 1993, with several (mostly recent) records for China (Carey et al 2001). In Thailand, the species is considered a vagrant to the central coast (Robson 2000).
The western African population is considered to be resident. The record mentioned before of an Italian ringed bird recovered in Chad is possibly the furthest inland record in Africa, although a few Saharan records are known for Algeria (Isenmann & Moali 2000), Morocco (Thévenot et al 2003) and Tunisia (Isenmann et al 2005). Other vagrants in Africa have been recorded in Ethiopia, Nigeria and Sudan (Urban et al 1986). In Kenya, it is considered a scarce migrant to the Rift Valley lakes, with a few coastal records (Zimmerman et al 1999). The species’ status in (landlocked) Uganda is comparable with inland Kenya, as it appears to be extending its East African wintering range (Carswell et al 2005). An exceptional record is that of an adult at Durban Bay, Kwazulu Natal, South Africa, on 10-13 September 1999 (Trevor Hardaker in litt).
In the Western Hemisphere, vagrancy to the Caribbean region has been suggested but could not be substantiated (cf Ebels 2002).
Discussion
Most populations in Europe have increased considerably over the last two decades, especially in the western Mediterranean. This period corresponds with the putative decrease of the Black Sea populations due to degradation of the wetlands on which the birds depended (Rudenko 1996). The population increases recorded in the western Mediterranean basin would therefore appear to derive from displaced birds from these populations (Sadoul 1997).
It is thought that some populations are not even viable without immigration. In southern France, for example, breeding success remained quite low while the population size has increased exponentially from the 1990s. The expansion, generally interpreted as an indication of a healthy population, does not match the observed breeding success (Sadoul et al 2003).
The recent upsurge of vagrant Slender-billed Gulls might be a consequence of conditions in the wintering range that could push some individuals to change migration routes. Hence, some birds decide to move northwards and others follow them (see Dispersal and immigration, Daniel Oro in litt). 
Increasing vagrancy in a northern direction has, in this species, also been explained as a response to climatic warming as with a number of other southern breeding species (Burton 1995). This last subject needs more research, as its effects on bird distribution are not yet fully understood. Increased observer awareness of the species’ subtle characteristics (Barthel & Königstedt 1993, Corso 1999) could be another factor but probably does not account for the rapid increase of records.
In order to have a further understanding of the geographic expansion of the Slender-billed Gull, and the ecological factors involved, a wider-scale metapopulation framework (including the eastern populations) is needed (Sadoul et al 2003).
Gert Ottens
Gert works for the BirdLife-partner “Vogelbescherming Nederland”.
Gert is also editor for Dutch Birding Magazine. www.dutchbirding.nl