Translator Disclaimer
1 April 2012 Revisiting the Proposed Leap-Frog Migration of Bar-Tailed Godwits along the East-Atlantic Flyway
Sjoerd Duijns, Joop Jukema, Bernard Spaans, Peter van Horssen, Theunis Piersma
Author Affiliations +

Two populations of Bar-tailed Godwits Limosa lapponica occur along the East-Atlantic Flyway. The European population (L. l. lapponica) is supposed to breed in northern Scandinavia and has been suggested to only winter in Europe. The Afro—Siberian population (taymyrensis) is supposed to breed in Northern Siberia and is thought to winter exclusively in West Africa. An analysis of 946 metal ring recoveries accumulated by EURING (with data going back to 1935), in combination with an analysis of over 13,000 resightings of almost 4000 individuals marked with colour-rings in 2001–2010, enabled us to examine whether there is evidence for overlap of the populations in summer and winter. Nearly all marked individuals behaved according to the previously suggested leap-frog migration pattern. On the basis of the present sample, only 0.8% of (colour) ringed birds that were recovered and/or resighted on the wintering grounds in Europe or West-Africa made a change between the two supposed wintering areas. This is far less than was previously estimated on the basis of biometric data. The distinct migratory behaviour of the two populations makes them near-completely separated in summer and winter. The Bar-tailed Godwit along the East-Atlantic Flyway thus exhibits a clear leap-frog migration, in which the Siberian breeders winter south of the European breeders.

Shorebirds provide excellent opportunities to study migration strategies. They occur in open landscapes and often rely on relatively few wetlands (Piersma 2007) where they can be captured, marked and resighted with relative ease (Piersma & Spaans 2004, Spaans et al. 2011). Not surprisingly then, shorebirds have their ‘connectivity’ well resolved (van de Kam et al. 2004, Delany et al. 2009).

In many migrating birds, populations breeding in northerly areas migrate to wintering areas south of populations from more southerly breeding ranges (Newton 2008). This so-called leap-frog migration occurs in several species of shorebirds (Salomonson 1955, Alerstam & Högstedt 1980, Alerstam 1990) and is thought to occur in Bar-tailed Godwits Limosa lapponica wintering in Europe and West-Africa (Drent & Piersma 1990, Scheiffarth 2001). The European population breeds and winters in Europe (breeding from Scandinavia to the Kanin peninsula and resides around the North Sea and Irish Sea in winter), and the Afro-Siberian population breeds in north-central Siberia (from Yamal peninsula in the west to the delta of Anabar river in the east) and winters along the west coast of Africa, with large concentrations on the Banc d'Arguin, Mauritania, and in Guinea-Bissau (Scott & Scheiffarth 2009).

Figure 1.

The seasonal itinerary of the two populations of Bartailed Godwits (European and Afro—Siberian) indicating the sequence of phases experienced by the two populations in the course of the year (from Drent & Piersma 1990).


The characterization of the leap-frog migration pattern (Drent & Piersma 1990) was based on Prokosch (1988), who found morphological differences in time and space suggestive of subspecific differentiation. Engelmoer & Roselaar (1998) proposed that the two Bar-tailed Godwit populations should be recognized as distinct subspecies. They named the birds with smaller body dimensions breeding in north-central Siberia taymyrensis and proposed to retain the larger-bodied European population as the nominate subspecies lapponica. When reviewing Engelmoer & Roselaar (1998), Tomkovich & Serra (1999) argued about some of their subspecies assignments, but not about the distinction between lapponica and taymyrensis. In later studies the two populations appeared to be not only morphologically, but also ecologically distinct (Scheiffarth et al. 2002, Duijns et al. 2009).

Based on morphological measurements of birds captured in the Wadden Sea, and discrimination functions based on museum specimens from the breeding grounds, Engelmoer (2008) estimated that about 20% of the Bar-tailed Godwits wintering in the Wadden Sea belong to the Afro-Siberian population. This implies that of the 120,000 birds wintering in the Wadden Sea (the European population; Scott & Scheiffarth 2009), no fewer than 24,000 individuals represent Afro—Siberian birds that were supposed to all winter in West-Africa. If so, this would mean that the leap-frog migration pattern is partial at best.

In this paper we aim to reconsider all available evidence using historical ringing, recovery and colourring resighting information of Bar-tailed Godwits along the East-Atlantic Flyway. Based on seasonal itineraries (Figure 1), we derived three criteria to assign individuals to either population at capture and ringing: (1) individuals (re)captured and/or re sighted between November and March in Europe belong to the European breeding population, (2) individuals captured and/or resighted in West-Africa belong to the Siberian breeding population, and (3) individuals captured in the Wadden Sea during autumn in active primary moult are also expected to belong to the European population, as wing-moulting individuals tend to winter in Europe (Atkinson 1996). We use then the recoveries and resighting data to establish whether the suggested leap-frog migration pattern of the two subspecies holds up.


Bar-tailed Godwits were captured at various sites in the Dutch and German Wadden Sea and on the Banc d'Arguin, Mauritania, West-Africa. Birds were processed immediately after capture, length of bill (exposed culmen, from tip of bill to base of feathers), wing (flattened and straightened; Prater et al. 1977) and tarsus being measured on most individuals using standard methods. The primary moult score was given according to Newton (1966); old: 0; growing: 1–4; new: 5. A bird that had completed moult of all 10 primaries had a primary moult score (PMS) of 50.

Each Bar-tailed Godwit was marked individually with four colour-rings (blue, red, white and yellow), combined with one yellow or red flag, and a metal ring. There were two colour-rings on the left and two on the right tarsus, and the metal ring was placed on one of the tibiae, but was not part of the code. The flag was the marker of the scheme and was placed on the tarsus or on one of eight different positions. In this way 2048 combinations were possible per flag colour. The colour-ring combination could easily be observed in the field by telescope. Unfortunately, mistakes were sometimes made as ring colours deteriorated through time (Burton 2000, Ward 2000). One should keep this in mind in the case of exceptional life-histories based on single ring-reading occasions. From spring 2001 to the end of 2010 a total of 3996 individuals were colour-ringed and 13,326 individual resightings from 2373 individual birds (59% of birds marked) were received from 311 different locations. The majority of the colour-ringed birds were caught in the Wadden Sea (91%), followed by Mauritania (7%). The colour-ring resightings show the same geographic bias, as most of the birds were resighted in the Wadden Sea (87%) followed by WestAfrica (11%). A similar pattern is observed for the metal rings. Most birds were captured in Western Europe (85%; i.e. United Kingdom, The Netherlands, and Germany), and recovered in Western Europe (85%; Appendix 1).

Figure 2.

Recoveries and resightings of Bar-tailed Godwits along the East-Atlantic Flyway. (A) Metal ring recoveries (n = 946) from 1935–2010; (B) colour-ring resightings (n = 13,326) from 2001–2010.


Figure 3.

Morphological characteristics of Bar-tailed Godwits, by sex and population. The box-and-whisker plots show median (line in box), interquartile range (box), range (bars), and outliers (small dots) of: (A) bill length (B) wing length and (C) tarsus length.


Table 1.

Assignment criteria of ringed and colour-marked Bar-tailed Godwits to the two populations and the verifications testing the leap-frog migration hypothesis based on resighting and recovery locations: the eight exceptions are listed in Appendix 2. The n values refer to the total number of individuals. Individuals may feature in different categories (e.g. an individual was caught and resighted), and therefore the totals differ from the sum of the separate categories.


From the EURING database 946 recoveries of metal rings were obtained, with the earliest recovery dating from 1935 and the latest from 2010. A preliminary analysis showed no spatial or temporal difference between earlier (< 1980) and later records, so all recoveries were used. In total 790 catching or recovery locations were identified. From only 35% of the individuals, relevant biometric (age and sex) information was available. To avoid reducing the sample size, all individuals were therefore included in the analysis.

Capture and resighting data were used to create a map with a resolution of 0.25 degrees (Figure 2). Resighting colour-ringed individuals is highly dependent on volunteers, and therefore the data were skewed towards locations where volunteers were active. To reduce the effect of identification errors, only individuals that were resighted twice in their wintering areas (i.e. West-Africa or Western Europe) were included in the analysis (n = 1399). Most Bar-tailed Godwits were caught and/or resighted during spring migration in the Dutch Wadden Sea in May when both populations occur in the Wadden Sea (Drent & Piersma 1990, Duijns et al. 2009), and therefore 790 (56%) of the colour-ringed birds could not be assigned to any population. Furthermore, only adult birds were included in the analysis as juvenile Bar-tailed Godwits may migrate differently to adults (B. Spaans et al. unpubl. data) and they are known to be scarce at Western European staging sites in spring (Prokosch 1988). This age-differential migration is not uncommon in migrating birds (e.g. Cristol et al. 1999, Lok et al. 2011).

To test for differences in morphological variables between the two populations, we performed an ANOVA with population and sex as fixed factors and date of catch as a covariate. Basic assumptions of parametric tests were examined by testing for normality with a Kolmogorov—Smirnov test, and the application of the Levene's test for equality of variances.


Despite a large overlap in morphological variables (Figure 3), bill and wing length (but not tarsus) confirmed that birds of the European population are larger-bodied than birds assigned to the Afro-Siberian population (ANOVA, F1,381 = 39.21, P < 0.001, F1,278 = 15.8, P < 0.001 and F1,372 = 0.5, P = 0.481, respectively).

Of the assigned individuals, 224 (16%) colour-ringed birds wintered in Western Europe, or were in active primary moult in autumn; 385 (28%) wintered in West-Africa. Of the 946 metal ring recoveries, 291 (31%) individuals wintered in Europe and 68 (7%) individuals wintered in West-Africa. Most of the marked individuals that were recovered or resighted behaved as predicted on the basis of the previously inferred leap-frog migration pattern (Table 1). As predicted, the two colour-ringed individuals that were observed in the breeding range in Northern Scandinavia were resighted in Europe in winter, and the 27 metal-ringed individuals assigned to the European population were recovered in Northern Scandinavia (Figure 4A). Of the metal-ringed birds, 23 African-winterers were recovered in the Northern Siberian breeding range (Figure 4B), thus confirming the links between wintering areas and breeding grounds. Of the 992 assigned birds (i.e. European or Afro-Siberian), only eight (0.8%) individuals did not follow the predictions. This included four colour-ringed birds and four metal-ringed birds (Appendix 2).


Due to the low density of breeding birds and the very low ring-reading efforts on the breeding grounds, we received only two resightings in the breeding areas. Yet, the recoveries from the Scandinavian and north-central Siberian areas support a leap-frog migration system, with little evidence for overlap of the breeding populations in winter. The leap-frog migration hypothesis is further supported by the observation that of 1009 birds caught in May and resighted more than once, only 3.8% were resighted in Europe during winter (Spaans et al., unpubl. data). Similarly, Wilson et al. (2007) found “low levels” of exchange for two other populations of Bar-tailed Godwit (menzbieri and baueri) with a comparable migration system. The eight exceptions (Appendix 2) were in fact all quite peculiar in terms of age (at the time of capture less than 2 years old), recovery dates (i.e. mid-April and mid-May when such individuals should still be in Europe), or ring colour (white and yellow may have been confused). Even if correct, these individuals switching wintering areas represent a small proportion of the population, and this suggests that the estimate by Engelmoer (2008) of 20% of the wintering population in the Wadden Sea as north-central Siberian-breeding (Afro—Siberian), is an overestimate. Our results thus suggest almost complete separation of the wintering and breeding grounds of the two populations of Bar-tailed Godwits along the EastAtlantic Flyway, and confirm that the two populations represent a clear example of a leap-frog migration system as Drent & Piersma (1990) suggested it to be.

Figure 4.

Wintering, staging, and breeding sites for Bar-tailed Godwits. (A) Recoveries of the European population with the main wintering sites in the Dutch and German Wadden Sea, the NW of Spain and the SE of the United Kingdom. (B) Recoveries of the Afro-Siberian population, with ‘hot spots’ in the Dutch and German Wadden Sea and the Banc d'Arguin in Mauritania.



We acknowledge the European Union for Bird Ringing (EURING), and all European ringing schemes in providing the recovery data and especially Henk van der Jeugd from the Dutch Centre for Avian Migration and Demography. We thank wilsterflapper Catharinus Monkel, the ringing groups VRS Castricum and VRS Calidris Schiermonnikoog for their considerable efforts to catch and colour-ring Bar-tailed Godwits. We also thank Jacintha van Dijk, Rob G. Bijlsma, Jouke Prop and Meinte Engelmoer for substantial feedback on earlier drafts. This study would not have been possible without the dedication of all other ringers and reporters who captured, ringed and reported recoveries world-wide. We especially thank Harry Horn, Laurens van Kooten and Jan de Jong for their amazing resighting efforts.



T. Alerstam 1990. Bird migration. Cambridge University Press, Cambridge. Google Scholar


T. Alerstam & G. Högstedt 1980. Spring predictability and leapfrog migration. Ornis Scand. 11: 196–200. Google Scholar


P.W. Atkinson 1996. The origins, moult, movements and changes in numbers of Bar-tailed Godwits Limosa lapponica on the Wash, England. Bird Study 43: 60–72. Google Scholar


N.H.K. Burton 2000. variation in sighting frequencies of colour-ringed Redshanks Tringa totanus according to ringing-scheme and ring colour. Wader Study Group Bull. 91: 21–24. Google Scholar


D.A. Cristol , M.B. Baker & C. Carbone 1999. Differential migration revisited: latitudinal segregation by age and sex class. Curr. Ornithol. 15: 33–88. Google Scholar


S. Delany , D. Scott , T. Dodman & D. Stroud (eds) 2009. An atlas of wader populations in Africa and Western Eurasia. Wetlands International, Wageningen, pp. 291–297. Google Scholar


R. Drent & T. Piersma 1990. An exploration of the energetics of leap-frog migration in arctic breeding waders. In: E. Gwinner (ed.) Bird migration, physiology and ecophysiology. Springer-Verlag, Berlin, pp. 399–412. Google Scholar


S. Duijns , J.G.B. van Dijk , B. Spaans , J. Jukema , W.F. de Boer & T. Piersma 2009. Foraging site selection of two subspecies of Bar-tailed Godwit Limosa lapponica; time minimizers accept greater predation danger than energy minimizers. Ardea 97: 51–59. Google Scholar


M. Engelmoer 2008. Breeding origins of wader populations utilizing the Dutch Wadden Sea, as deduced from body dimensions, body mass, and primary moult. PhD thesis, University of Groningen. Google Scholar


M. Engelmoer & C.S. Roselaar 1998. Geographical variation in waders. Kluwer, Dordrecht. Google Scholar


T. Lok , O. Overdijk , J.M. Tinbergen & T. Piersma 2011. The paradox of spoonbill migration: most birds travel to where survival rates are lowest. Anim. Behav. 82: 837–844. Google Scholar


I. Newton 1966. The moult of the Bullfinch Pyrrhula pyrrhula. Ibis 108: 41–67. Google Scholar


I. Newton 2008. The migration ecology of birds. Academic Press, London. Google Scholar


T. Piersma 2007. Using the power of comparison to explain habitat use and migration strategies of shorebirds worldwide. J. Ornithol. 148 (Suppl. 1): 45–59. Google Scholar


T. Piersma & B. Spaans 2004. The power of comparison: ecological studies on waders worldwide. Limosa 77: 43–54. Google Scholar


A.J. Prater , J.H. Marchant & J. Vuorinen 1977. Guide to the identification and ageing of holarctic waders. British Trust for Ornithology, Tring. Google Scholar


P. Prokosch 1988. The Schleswig-Holstein Wadden Sea as spring staging area for arctic wader populations demonstrated by Grey Plover (Pluvialis squatarola, L. 1758), Knot (Calidris canutus, L. 1758) and Bar-tailed Godwit (Limosa lapponica L. 1758). Corax 12: 273–442. Google Scholar


F. Salomonsen 1955. The evolutionary significance of bird migration. Dan. Biol. Medd. 22: 1–61. Google Scholar


G. Scheiffarth 2001. Bar-tailed Godwits (Limosa lapponica) in the Sylt-Rømø Wadden Sea: which birds, when, from where, and where to? Vogelwarte 41: 53–69. Google Scholar


G. Scheiffarth , S. Wahls , C. Ketzenberg & K.M. Exo 2002. Spring migration strategies of two populations of Bar-tailed Godwits, Limosa lapponica, in the Wadden Sea: time minimizers or energy minimizers? Oikos 96: 346–354. Google Scholar


D. Scott & G. Scheiffarth 2009. The Bar-tailed Godwit. In: S. Delany , D. Scott , T. Dodman & D. Stroud (eds) An atlas of wader populations in Africa and Western Eurasia. Wetlands International, Wageningen, pp. 291–297. Google Scholar


B. Spaans , L. van Kooten , J. Cremer , J. Leyrer & T. Piersma 2011. Densities of individually marked migrants away from the marking site to estimate population sizes: a test with three wader populations. Bird Study 58: 130–140. Google Scholar


P.S. Tomkovich & L. Serra 1999. Morphometrics and prediction of breeding origin in some Holarctic waders. Ardea 87: 289–300. Google Scholar


J. van de Kam , B.J. Ens , T. Piersma & L. Zwarts 2004. Shorebirds, an illustrated behavioural ecology. KNNV Publishers, Utrecht. Google Scholar


R.M. Ward 2000. Darvic colour-rings for shorebird studies: manufacture, application and durability. Wader Study Group Bull. 91: 31–34. Google Scholar


J.R. Wilson , S. Nebel & C.D.T. Minton 2007. Migration ecology and morphometrics of two Bar-tailed Godwit populations in Australia. Emu 107: 262–274. Google Scholar


Er komen langs de Oost-Atlantische trekroute twee verschillende populaties van de Rosse Grutto Limosa lapponica voor. Een ‘Europese’ populatie (ondersoort lapponica), die in het noorden van Scandinavië broedt en naar wordt aangenomen uitsluitend in Europa overwintert. En een ‘Afro—Siberische’ populatie (ondersoort taymyrensis), die in Noord-Siberië broedt en verondersteld wordt de winter in West-Afrika door te brengen. Als we dit tot op heden theoretische onderscheid kunnen onderbouwen, dan hebben we een mooi voorbeeld van een ‘leap-frog’ migratiesysteem, waarbij de Afro—Siberische populatie ‘over’ de Europese populatie heen trekt. Wij analyseerden 946 terugmeldingen van metalen ringen opgebouwd door EURING (met gegevens die teruggaan tot 1935) en meer dan 13.000 waarnemingen van het NIOZ kleurringprogramma uit de afgelopen jaren (2001–2010). Deze analyse stelt ons in Staat om te onderzoeken of de Rosse Grutto inderdaad een ‘leap-frog’ migratie laat zien of dat er tussen beide overwinteringsgebieden uitwisseling plaatsvindt. Bijna alle gemerkte Individuen gedroegen zich volgens de gangbare theorie. Slechts acht vogels (0,8%) van de (kleur)ringmeldingen bleken beide overwinteringsgebieden (Europa en Afrika) gebruikt te hebben. Zij vertegenwoordigen dus opmerkelijke uitzonderingen. We kunnen zelfs niet met 100% zekerheid zeggen dat deze vogels inderdaad in beide overwinteringsgebieden geweest zijn, want mogelijk is er sprake geweest van verkleuring van de kleurringen. De mate van overlap in overwinteringsgebied tussen beide broedpopulaties is veel kleiner dan aanvankehjk op grond van de biometrie van de vogels werd gedacht. De verschollen in morfologie tussen de twee populaties zijn echter te klein om individuele vogels in het overwinteringsgebied aan een van de twee broedpopulaties toe te rekenen. De verschillen in trekstrategie en overwinteringsgebied tussen de Europese en Siberische broedpopulatie laten zien dat de populaties ook buiten de broedtijd bijna volledig gescheiden zijn en dat er inderdaad sprake is van een ‘leap-frog’ migratie.

Appendix 1

Appendix 1.

Total number of ringed Bar-tailed Godwits resighted and recovered per country.


Appendix 2

Appendix 2.

(Colour-) ringed individual Bar-tailed Godwits that were uiifaidiful to their wintering area, or that were reported in Europe in winter where they were expected to be in West-Africa at the time of resighting. If a bird was resighted twice on the same day, this means that the individual was observed by at least two independent observers.

Sjoerd Duijns, Joop Jukema, Bernard Spaans, Peter van Horssen, and Theunis Piersma "Revisiting the Proposed Leap-Frog Migration of Bar-Tailed Godwits along the East-Atlantic Flyway," Ardea 100(1), 37-43, (1 April 2012).
Received: 16 June 2011; Accepted: 1 January 2012; Published: 1 April 2012

Back to Top