Migratory progress of juvenile and adult Dunlin Calidris alpina from the Baltic and the Waddensea perspective.

by Christer Persson

Christer Persson, Ljungsätersvägen 43, S236 41
Höllviken, Sweden. E-mail cp.hollviken'at'swipnet.se

Straight text version for printing: text version.

Abstract. Many errors of Waddensea-based papers on Dunlin are caused by defective and misleading quotation technique where Baltic conditions are concerned, the very grave case of Goede, Nieboer & Zegers 1990 is exposed.
Another problem is connected with disregard or ignorance of facts, without explicit misquotation, Dierschke 1996 and Rösner 1990, 1997 represent this type of corruption; both when overlooking or disregarding the joint migration of juveniles and late adults in the Baltic, Rösner when overlooking or disregarding the swift migration of August juveniles, the staging of November juveniles and the joining of wintering adults and juveniles. These events are important to the overall assessment of juvenile progress. In addition Rösner's 1990, 1997 repeated quotation of mean values for juvenile stay in the Baltic is misleading, but he may have been unaware of the dangers. Finally, Rösner is wrong when discussing moult, he believes in extensive "moult on migration". Here his source is in error. The misleading source is primarily Holmgren et al. 1993b, these authors in some sense have wasted an inadequate field material from Ottenby by treating it from an inadequate theoretical point of departure; in their zeal to exhibit a clear-cut "adaptation" they were blind or blinded to the true extent of suspension, and there are no birds with fresh remiges in the material, although at least 5 and possibly 10 % of all Dunlin passing the Baltic area on migration have moulted completely on breeding-grounds.
Based partly on this foundation of ignorance and disregard, partly on correct information Dierschke and Rösner have developed general notions of Dunlin progress within and outside the Waddensea area. This paper attempts to correct the unsound foundation of their efforts, first by outlining juvenile and adult behaviour in a wind flat environment like the Baltic. Second it diversifies the moult issue by separating 2c and 3c+ birds; ostensible "moult on migration" is a pronounced feature in 2c birds, and by all likelihood accentuated from western to eastern populations; 75 % of 2c Dunlin with inner primary ranking 1 - 3 had initiated moult and suspended. 2c Dunlin of eastern provenience with hampering moult gaps may be overrepresented among resting birds during the main migration period of the Baltic. Third, a case of suspension in North Scandinavian Greenfinches Carduelis chloris is described in some detail; remiges don't grow although they are not completed, the gap distribution is much the same as in Dunlin, and gaps are clumped to the zero end (= small gaps). Up till now it has been tacitly assumed, that suspension implies a halt between a full-grown, new remige and a worn, old one. In Dunlin, it seems as if the growth of a remige may be suspended at any time, however, and not resumed again until the energetic demands from migration flight are mitigated or over with, such suspension has been interpreted as "moult on migration".
Finally, for Waddensea quotation, the following three facts are stated: 1. In the Baltic juvenile Dunlin arrive accompanied by adults throughout the autumn season. Adult guidance may be of importance in a wind flat area, with stochastic change of water levels, and adults know the way and the goal areas. 2. The overall behaviour of juveniles in the Baltic is purposeful and systematic, they are in no way "opportunistic", and evidence for such behaviour in the Waddensea should be reconsidered. 3. There is no factual material showing that adult Dunlin moult on migration, the distribution of gaps is more in line with suspended growth even of uncompleted remiges. These were shed and grown on or near breeding-grounds and growth will be resumed when the migration journey is coming to an end.


1.1. Introduction I: A hypothesis concerning the human aspect of ornithology
1.2. Introduction II: Movements of adult and juvenile Dunlin, in quotation
2. The Waddensea versus the Baltic: juveniles
2.1. Juvenile events in the Baltic, in correct quotation
3. The Waddensea versus the Baltic: adults
3.1. Adult events in the Baltic: female and male performance
3.2. Adult events in the Baltic: 2c birds and "moult on migration"
3.3. Can growth be suspended before a remige is full-grown and then resumed again?
3.4. Excurse: The Waddensea as a reserve of metaphysical indeterminism TO BE CONTINUED FROM HERE
4. Discussion
5. A collection of statements from the Baltic: adults and juvenile progress

1.1. Introduction I: A hypothesis concerning the human aspect of ornithology.

No single worker can embrace the whole life-cycle of Dunlins; he will depend on other workers to know those parts he cannot see for himself. This dependency, the need to quote the results and views of others, is the critical part of any Dunlin paper. Errors begin here. And not necessarily with the original statement, but with the emphasis, the bias of the quotation.

In the 1990's the end station of Dunlin migration, the Waddensea area, has been the centre of gravity of Dunlin study (Dierschke 1996, Goede, Nieboer & Zegers 1990, Rösner 1990, 1997), and other areas - their time schedules, their moult patterns - have appeared in quotation:

The Siberian juveniles leave the breeding area shortly after the departure of the adults. (...) In Scandinavia, these juveniles are already caught within two weeks after the passage of the first adults and a few weeks before the passage of the European juveniles. In consonance, Brenning 1987 noted two waves of migrating juveniles at Wismar (DDR) in autumn; the first small wave ocurred in the last decade of August, a second maximum was reached in the second half of September, most often in the last decade. (Goede, Nieboer & Zegers 1990)
But what does Brenning 1987 actually say? Here follows his text in English translation:
In most years the migration of juveniles begins in the last decade of August, often already by mid-August. A first maximum can be reached already by the turn August/September (up to 240 birds/day), but it is not always distinct or it will fall in the first September decade. The first more extensive wave of juvenile Dunlin reaches Langenwerder in the second half of September (not before 18.9), mostly in the last decade, in many years not until the turn September/October, when up to 500 birds/day may be counted. After a more or less distinct decline numbers rise again by mid-October (150 to 500 birds/day), and this maximum rarely fails to come. (Brenning 1987)
So, if Goede et consortes had turned the page, they would have discovered, that Brenning explicitly discerned three migration waves, that the second and the third were the major ones, and that the final wave might last into November. Furthermore, if they had paid some attention to the catching seasons of Swedish and Norwegian catching stations and compared them to Brenning's statements, they would have realized that no conclusions could be drawn from their own Waddensea material since so few juveniles were - and are - caught in Sweden or Norway between mid-September and early November, the main migration period of juveniles in Brenning's diagram. The evidence for an early passage of Siberian juveniles, dating a.o. from this paper, is essentially an artefact, produced by Dutch quotation technique. The early occurrence of Siberian juveniles together with "adult buff" adults in the last stage of moult (indicated by low, but not significant score value of Fig. 16 in "Phenology and biometry...") is a mere trickle compared to the massive presence in early November, and we do not know for certain, whether the wing-length and bill-length peak from the first half of September (Figs. 7 and 8 of "Phenology and biometry...") should be attributed to Siberian birds. In support of my criticism, I could add a quotation from one of the authors representing Baltic standpoints, Meltofte 1993; I do not think his remark is entirely to the point (the statement on culmination is incorrect, late juveniles are almost entirely night migrants, and there is no catch that can make up for the absence of observations), but at least it shows that there is no support for the Dutch view in one of the best informed and most comprehensive papers on Baltic conditions:
It is not possible to separate the migration of European and Siberian birds, but as the migration culminates c1 week earlier in SW Finland, at Ottenby and S. Amager (Copenhagen, Sealand) than at Blåvand (W Jutland) (Fig. 23; Edelstam 1972, Kaukola & Lilja 1972), it is probable, that North European breeding birds migrate a little earlier than Siberian.

This introduction could be repeated in countless versions, all highlighting the same problems: (in-)correct quotation, (ir-)relevance of results. Alienation from close neighbours was the cause, maybe mixed with a little condescendence: these days saw the repercussions of political division, Cold War. Goede, Nieboer & Zegers 1990 was published only a few months after the Berlin Wall had fallen, but it quoted in the alienated spirit of before-November 1989. The West Germans Dierschke and Rösner were already active by the same time; Dierschke in connection with Vogelwarte Helgoland, Rösner working with a WWF-financed project in the Waddensea. They came from an environment of basic German alienation, a stance facing the Atlantic and turning its back even on the Baltic. Only a few years later tensions were beginning to dissolve; Dierschke was a pioneer in testing the good-will of his fellow countrymen: first at Langenwerder, later on at Hiddensee. But did he "open his mind" to the new environment, or was he, like so many West Germans, just bringing West Germany (here: the Waddensea perspective) to East Germany (the Baltic) without learning? This is my hypothesis: it takes a new generation to "look east" where Dunlins are involved, the Waddensea generation - Dutch or German - only sees what it was brought up to see, and due to the specific conservatism of biology, its hierarchic and largely un-checked quotation structure, such errors may prevail if they are not exposed and opposed. Dierschke's dogmatic "differential migration" (Dierschke 1996) and Rösner's "opportunistic juveniles" and "conservative adults" (Rösner 1997) are Waddensea constructs. In their cases wrong quotation is not the main problem, but rather the fact that they quote Waddensea when they should quote something else: the Baltic or Russia or Siberia. Rösner's alleged "synthesis" ("A conceptual model of the migratory system of Dunlin"; Rösner 1997) is a thesis, no more, and in order to oppose it I will have to draw up this paper as an antithesis; I must quote myself in deliberate opposition. (This promises to be quite a marvelous Hegelian Auseinandersetzung). And whatever synthesis will follow, it will be about correct quotation: how should the fact that juveniles are accompanied by adults in the Baltic be quoted in the Waddensea, how should the fact that juveniles are very goal-oriented and purposeful in the Baltic be quoted in the Waddensea, how should the fact that moulting migrants suspend remige growth be quoted in the Baltic and the Waddensea?

Introduction II: Movements of adult and juvenile Dunlin, in quotation.

There is a general consensus that a majority of adult Dunlin stick to traditional migration routes as well as moulting and wintering grounds (Brenning 1989, Fuchs 1973, Gromadzka 1989, Hardy & Minton 1980, Jönsson 1986, Meltofte 1993, Minton 1975, Pienkowski & Dick 1975, Pienkowski 1976, Roos 1984, Rösner 1990, 1997). Exceptions from the rule have been attributed to a possible "regular deviation" (Nørrevang 1955) of adult migration or deterioration of habitats and unfavourable weather (Hardy & Minton 1980, Rösner 1990, 1997). In contrast juveniles are said to disperse more (maps of recoveries), rest longer (average stay) and be more liable to test unfamiliar areas and possibilities (Brenning 1987, Gromadzka 1983, 1989, Holmgren et al. 1993a, Pienkowski & Evans 1985). But which juveniles linger? According to Brenning 1989 the early (August) juveniles disappear very quickly, and there are no signs of them in Europe in winter, the same conclusion may be drawn from the maps in Gromadzka 1983, 1989. Some juvenile Dunlin linger in the Baltic, some leave it very quickly, with extreme migration speed. From this unclear state, alluding to some average state, quoting mean values, Rösner 1990 elaborated a kind of ideology for juvenile progress:
Before their first migration juveniles have only a vague concept of direction to and location of resting and wintering grounds. They detect them by applying the method "trial and error" and visit the favourable areas anew in years to come.
This in turn should explain why juveniles should be more likely to detect new, favourable possibilities, and why they should be quicker to abandon decaying areas. Dierschke 1996 makes a similar kind of emphasis already in his title: "Differential migration". Here is a very strong simultaneous statement by two noted Waddensea workers; I do not underestimate the strength of defence they are liable to adduce in favour of their cause. But the arguments so far brought forth are not convincing, nor is the quotation pertaining to Baltic conditions. I could add: the presupposed independent progress of juveniles creates secondary problems, e.g.: how can breeding populations retain some sort of genetic homogeneity when birds from all areas mix indiscriminately in the Waddensea? Wenink and Baker 1994 note this "confluence" of populations on southward migration, their remarks gathered up by Rösner 1997 in support of his own position, but do they not imply another complication, another questioning of his views?

2. The Waddensea versus the Baltic: Juveniles

2.1. Juvenile events in the Baltic, in correct quotation. Three "events" connected with juvenile Dunlin migration in the Sound area are disturbing to Dierschke's and Rösner's general approach: 1) the quick disappearance of August juveniles from the Baltic, 2) the staging of juveniles in October / November and 3) the organization of Baltic wintering from November onwards. All cases presuppose navigational ability and knowledge of goal areas, either in juveniles or at least among adults accompanying them; confronted with these facts we must ask: how long are age-classes differentiated and where does the separation expire? As far as I can see, the differences are exaggerated by both Dierschke and Rösner. The consistency of Rösner's spring bill measurements from the Waddensea is noted in "Bill length distributions..."; but the same kind of consistency is characteristic of juvenile migration from the Baltic or Sound perspective; in particular I wish to call attention to Figs. 14 and 16 of "Phenology and biometry...". There is a continuous development throughout the autumn, indicating a constant, almost methodical turnover, and this development is broken only by sudden, discontinuous "jumps" (biometry of Figs. 7, 8 and 9 of "Phenology and biometry..."). This pattern doesn't call "trial and error" to mind, nor does the elaborate mechanism that brings wintering juveniles and adults together. (I do not expect either Dierschke or Rösner to advocate trial and error in the choice of these juveniles; I suppose they agree that wintering juveniles in Öresund are progeny of the adults they are joining). Against this background I question the notion of random dispersal, even in the Waddensea: each single juvenile Dunlin has a goal, its way of getting there, its own, characteristic progress. I will elaborate my views in two steps, with some extra supporting material:

2.1.1. Juvenile Dunlin in the "shifty" Baltic.. Feeding-grounds of the Baltic, the Belt Sea and the Sound could never really be termed favourable or dependable; in some years there will be practically no wrack beds at Ottenby, on the Falsterbo peninsula or at Bottsand (Bay of Kiel), and wind flats will remain submerged for much of the migration period. Under such circumstances the whole area must appear as an "error" to inexperienced juveniles, and all attempts at wintering should be interrupted. (Dierschke 1996 is conscious of this problem, but like Rösner he believes or claims, that juveniles migrate indepently of adults). Later on, when autumn storms begin to blow, things may get worse still: water levels will inevitably rise with low air pressure and winds from NW, and even meadows bordering the coast may get flooded. With high water all Charadridae and Scolopacidae - including e.g. Grey Plovers Pluvialis squatarola - are forced to enter pasture or meadows above the waterline. Adults react promptly to this need, often involving a measure of niche expansion; in many cases they will have immediate experience of wind-flats from three areas: The Baltic, the Mediterranean and the the Black Sea (van der Have & van den Berk 1994). To juveniles, on the other hand, the situation is unfamiliar, without precedence, and there will always be some delay before they follow. (Still, they do not abandon the area, they follow).

Normally meadows or pasture are visited only by Snipe Gallinago gallinago and Jacksnipe Lymnocryptes minimus at this time of year; to Dunlin they do obviously not match the wind flats and will remain unvisited as long as the latter are exposed - at least in autumn. Meadow / pasture will serve for subsistence, no more. My point is: there is an important element of learning or following in waders; in most cases juveniles do not "detect", instead opportunities are pointed out to them by adults (in this way juvenile Golden Plovers Pluvialis apricaria and Ruff Philomachus pugnax "detect" farmland, juvenile Whimbrel Numenius phaeopus "detect" the location of Empetrum nigrum in Baltic staging areas, juvenile Greenshanks Tringa nebularia "detect" confined fish when brackish marshland dries up, and juvenile Sandpipers "detect" pasture when wind flats get flooded). The protracted "apprentice" period of young Oystercatchers Haematopus ostralegus is a well-known fact, but there is only a difference in degree between this specialized species and other wader species; the latter benefit from adult experience as well. And the Zugordnung of many wader species involves some degree of highly instrumental adult delay, that will mix adults and juveniles - of particular importance in the Baltic, where inexperienced juveniles are confronted with stochastic change of water levels for the first time in their lives. In any situation involving scarcity or stress in the Sound area - and most probably everywhere else as well - adults are relatively overrepresented where alternative possibilities are exploited. Three to five adults in a hundred juveniles will suffice - and they are present till mid-October in the case of Dunlins. The myth of juveniles as enterprising pioneers has a touch of vitalist ideology to it; in practice juveniles are not equally handy and less quick to react to situations of scarcity, and learning seems to be more important to them than "trial and error". The difference between age classes at the outset of migration is soon outbalanced by a community of interest and a similarity or even mimicry of behaviour.

2.1.2. Juveniles - whether lingering or migrating - behave purposefully. Juvenile Dunlin are brought to the Sound area by prominent, converging guiding lines (the shore-lines of Kattegat, the Swedish west coast, the south coast of Scania, the east coast of Sealand). Many juveniles wintering in the North Sea area and the Irish Sea are likely to have passed this area on their first migration. Still only a fraction of the overall passage stays behind, and one group of the birds present puts on large amounts of fat and continues to a probably very distant wintering area along the "Mediterranean Flyway" (Gromadzka 1989, van der Have & van den Berk 1994). Another cohort has come to a halt and will move no more before the following spring. In both these cases the presence of simple, but imperative "instructions", aiming at bringing birds together, is felt very strongly; a "trial and error" process could not possibly achieve the same end. Juveniles are to join their own kin (group, population) before some deadline, at the latest in the staging area in spring, I believe this encounter to take place much earlier; juveniles with medians tending to "adult buff" and with centralis biometry put on some 15 grams of fat and join their kin in - say - the Mediterranean in November, another contingent of juveniles with more pronounced alpina morphology meets with their kin in the Sound area from November onwards. Both cases presuppose some degree of navigational ability among juveniles, but in September and October adult guidance may be involved as well. The following two recoveries could relate to juveniles trying their luck (and discovering their error) by flying in opposite directions, but are they not more likely to be latitudinal and/or longitudinal corrections by birds with fully developed navigational ability, or associated with such birds?

  • 1c, Revtangen, Norway 30.8.97 - contr. Skanör 12.9.97. 576 km SE, 13 days
  • 1c, Langenwerder, Germany 5.10.99 - contr. Skanör 6.10.99. 178 km NNE, 10 hours

Rösner 1990, 1997 discusses this possibility in detail, to some extent repudiating Meltofte 1993 and explicitely stating: things could barely function this way - in contrast I think that they do. Since birds from such a vast area converge in Western Europe, there must be powerful mechanisms aiming at encounter and separation in the "confluence" of populations (Wenink & Baker 1994); any other alternative, e.g. one involving consequent "trial and error" on the part of juveniles would bring about indiscriminate gene flow and scatter juveniles all over the breeding distribution. There are two, very swift movements of juveniles from the Sound area early in the season, these early birds (cf. Brenning 1989) seem to know exactly where they are heading, and they do not stop for the German or Dutch Waddensea. (There are similar recoveries from other Swedish dating back to times when Dunlin were more hunted, e.g. 1c, Ledskär 10.8.60 - Ile aux Oiseaux, Gironde 20.8.60; 2155 km SW, 10 days):

  • 1c, Skanör 17.8.02. - contr. Lagoa Santo André (38.05 N, 8.47 W), Setubal, PORTUGAL 6.9.02. 2515 km SW, 20 days.
  • 1c, Skanör 17.9.97. - contr. Admiralty Point (52.48 N, 00.20 E), Terrington, Norfolk, ENGLAND 20.9.97. 864 km WSW, 3 days.

There are no really corresponding recoveries in Roos 1984, the swiftest ones are: 1c, Skanör 13.8.73 - Norfolk 2.9.73 and 1c, Skanör 6.9.73 - Suffolk 25.9.73, times elapsed 19 and 20 days. (My catching routines may be the reason; when the catch is moderate I try to handle the birds within minutes after the catch, avoiding bags, measuring and ringing directly from nets). But if recoveries from all north European schemes are pooled there is a lot of similar evidence over years, a joint treatment of these recoveries appears as highly desirable. I can only state as my conviction, based on the totality of experience from my field-work in the Sound area, that Dunlin juveniles navigate at a very early age, from September onwards, (or that they associate themselves with navigating adults), and that they have already reached some sort of goal, a staging area, an encounter area, when the researcher notices their apparent random movement or random distribution. And: I am convinced that the Sound pattern of encounter and separation is repeated from Orkney to Mauritania and the Mediterranean at roughly the same time of year, masked by the seeming chaos of the main wintering areas. This chaos makes Rösner resort to a stochastic thought model and reify it into "reality" - for obvious reasons, but incorrect. The position of the Sound at the outskirts of the main moulting and wintering area allows us to see what is really taking place; here lies the main advantage with field studies in such an area.

3. The Waddensea versus the Baltic: Adults

3.1. Adult events in the Baltic: female and male performance on migration

3.2. Adult events in the Baltic: 2c birds and "moult on migration". Rösner 1997 discusses the moult topic in detail, but his quotation pool is deplorably limited, quoted are mainly: Gromadzka 1986, Holmgren et al. 1993b. The most urgent task for Baltic workers today must be to expand the factual material connected with moult; the alleged "moult on migration" should be illuminated from all sides, in every possible way. For a start 3c+ and 2c birds should be separated as far as possible; there is an urgent need to exchange worn remiges primarily in 2c birds. 60% of all "moult on migration" in the Baltic can be attributed to 2c birds, although their share of the total breeding population is only 35 - 40 %, after years of poor reproduction substantially less.

Between 11 July and 30 August 1995 - 2004 there are 1367 age-determined adults from the Öresund area with known moult status, 1127 of these also ranked according to the Meissner scale: 891 aged 3c+, 476 aged 2c. 60.7 % of all summer adults were ranked 0, 39.3 % 1 - 3, i.e. the mean 0 ratio falls well below the mid-August starting value (c80 %) of juveniles in Figs. 16 of "Phenology and biometry...", indicating a clear eastern bias in the main body of adults migrating by way of the south Baltic in summer. In Fig. 1 this material is cross-tabulated from four aspects: age versus ranking, age versus moult, 3c+ ranking versus moult and 2c ranking versus moult. Rankings 1 - 3 were pooled into one class.

A G-test of ranking against age gives unadjusted Gadj = 0.36; there is no cause to reject the null-hypothesis that ranking is independent of age in summer adults. 2c and 3c+ birds are drawn from the same overall population. Second, a G-test of moult against age gives unadjusted G = 262.6, the null-hypothesis that moult is independent of age can be rejected at the 0.001 % level. Third, moult depends on ranking in both 3c+ birds (Gadj = 10.72, p < 0.01) and 2c birds, although the certainty of rejection is less here (Gadj = 6.24, p < 0.05).

In spite of the fact, that the overall passages of 3c+ and 2c birds are most likely drawn from the same pool, there are differences in the ranking patterns when the analysis goes into detail. In both categories "0" birds dominate in pentades 15.7 - 19.7 and 20.7 - 24.7, but after that - the main migration interval of the Baltic - there follow three pentades with 60 - 70 % ranking 0 in 3c+, but only 50 % ranking 0 in 2c. The material is presented in Table I:

Table I. Inner primary rankings of 3c+ and 2c Dunlin in the Sound area 25.7 - 8.8 1995 - 2002.

AgeRanking 0Ranking 1 - 3% ranking 0;
binom. s.d.
3c+1337464.3; 3.3
2c635752.5; 4.6

Both np and n(100-p) > 500, so the difference between percentages may be tested with normal distribution procedures; Z = 1.99, i.e. the difference is significant at the 5 % level. Similarly a G-test gives unadjusted G = 4.37; the null hypothesis that ranking is independent of age in this particular 15-day-interval can be rejected at the 5 % confidence level. From Table II of "Migrating Dunlin in the Baltic area: the moult issue" it is clear, that all migrating Dunlin, whether moulting or not, have much the same average weights; no category of summer migrants in the Baltic overexerts itself. The reason for the concentration of ranking 1 - 3 2c Dunlin with high frequency of initiated moult to the main migration period of the Baltic - accentuating the impression of a unique "moult on migration" - must be some other than low weights. But average weights do not imply average fat depots; Dunlin may achieve average weight in a few minutes simply by drinking or gathering up a few grams of food of low caloric content. Since all catches were made in the nighttime in the Sound area, many weights were taken just after a bird had descended from migration; there is no other field material of similar representativity in this respect. The matter should be investigated from the angle of fat content, however, not weights; I still believe that 2c moulters rest to a higher extent than other categories (even than unmoulted 2c birds, possibly with some other average ranking).

Turning to 3c+ birds there are ranking 1-3 peaks before and after the main period, indicating that 3c+ birds have a specific time schedule; there is no overall "adult" pattern. The lowest values for ranking 0 in 3c+ birds fall in pentades 20 - 24.7 (57.7 %; 5.3) and 9.8 - 13.8 (44.8 %; 6.5 - the main "adult buff" interval), indicating two peaks of older birds with ranking 1 - 3 (and possible eastern origin). In addition there is by all likelihood a third (female) peak in late June/early July. Here again the advantage of age and experience in Dunlin comes to the front; 3c+ birds have the most diversified schedule. This section attempts to indicate the possibilities of inner primary ranking, but the materials should be larger: 5000 ranked and sexed adults between 15 July and 15 September are needed for a conclusive analysis of the oscillatory character of adult migration. The main suggestion of the above contents is: Moulting 2c birds of eastern origin may be "selected" for the ringing stations in the Baltic, thereby creating a biased picture of the extent of moult in migrant Dunlin. This hypothesis may be tested in several ways, one is to classify in detail the situation (headwind, rain, general character of migration) out of which each single bird was taken.

3.3. Can growth be suspended before a remige is full-grown and then resumed again? In the section "Wing moult and migration" Stresemann & Stresemann 1966 give prominence to the fact, that only one migrant in more than 1000 North American tower and mast kills had growing primaries; the authors, Tordoff & Mengel 1956 (link not active), conclude that moult influences the migration period of an individual. To this Stresemann notes, with a philosophical "Umkehrung": This is true to a certain extent, but there is more general validity to the sentence: Migration has an influence on moult.

There is no way of knowing momentarily if an incomplete remige is still growing or if its growth has been "suspended". The bird must be held captive, fed and studied anew at significant intervals. Still, the suspicion that artificial conditions influence the result cannot be avoided, this must have been one reason why Holmgren et al. 1993b quoted negligible growth in captive Dunlin as: some feathers were not of full length and were found growing in a sample of caged individuals - they didn't trust their own results. In autumn I have seen Great Woodpeckers Dendrocopos major and Tengmalm's Owls Aegolius funereus migrating with incomplete remiges under conditions, that strongly suggested that these feathers were not growing at all, but I have no evidence in support of my view. In another species there is evidence, however: the Greenfinch, Carduelis chloris. This species has colonized North Scandinavia in the second half of the 20th century, the relatively late breeding and the protracted individual moulting period (Finland 75 - 85 days, Ginn & Melville 1983) keep northern adults from migrating until mid-October. When these populations occur in South Scandinavia on migration (late October - early December) they are always characterized by a high ratio of birds with "growing" outer primaries and inner secondaries, in addition the moult of 1 - 4 inner secondaries may be suspended or arrested. Their northern provenience is shown by numerous recoveries and controls. In Scania they may rest for a couple of days, particularly in rainy weather, and all retraps at feeding-tables have shown that these remiges do not grow at all (less than accuracy of measurements in 24 - 48 hours) as long as the birds are migrating, the completion must be deferred until the final goal area is reached. In order to establish this possibility, this apparent "moult on migration", the time distribution of gaps in both sexes and the overall frequency distribution of gaps is shown in Figs. 2, 3 and 4. The material was collected at Ljunghusen on the Falsterbo peninsula between 1979 and 2001.

The overall picture is strongly reminiscent of Fig. 8 and 9 in "Migrating Dunlin in the Baltic area: the moult issue", the clumping to the zero end is much the same (coefficient of dispersion of gaps 0.1 - 0.9 1.73, of overall material 3.59); most birds do not migrate until the average gap is 0.4 - 0.5 remiges, mainly in terms of secondaries - assuming that they leave N. Scandinavia with much the same gap. It is still unclear when effective feather growth is resumed, in some cases it never happens, the birds continue to fly with a gap till next moult, which is clear from the following moult card; the bird had never grown a suspended S3 to full length. In spring I have not noted a frequency of gaps corresponding to that in autumn, indicating that the birds do resume growth, on the other hand I probably sample more southerly populations in spring than in autumn. The question of resumed growth should be solved in the N Scandinavian breeding area: are inner secondaries full-grown here in spring?
  1. Greenfinch, 2c+ female, Ljunghusen 21.8.89. P1 - 4 new, P5 shed; S1 - 2 "semi-new", of full length, S3 "semi-new " 0.9, S4 - 6 older and very worn.

This excurse serves to establish the fact, that birds are capable of suspending before a remige is full-grown and of resuming feather-growth at some stage, by all likelihood when the migration journey is over. The average gap of Dunlin in the Sound area (0.3 - 0.4) and Greenfinches (0.45) is much the same, and I think that far-reaching conclusions are permitted; birds - and in this particular case, Dunlin - are capable of suspending when a feather is not full-grown and growing on "low jets" or resuming growth at a later stage. Next, the growth rates of pins (length 0.1 - 0.3) and of practically full-grown remiges (0.7 - 0.9) should be compared: is suspension equally effective in both cases?

3.4. Excurse: The Waddensea as a reserve of metaphysical indeterminism. If we choose to describe the Brownian movement of particles in a gas volume by means of e.g statistical mechanics, this doesn't necessarily imply an assumption, that the world per se (or only the gas volume) is stochastical in some ontological sense. There is still cause and effect behind the movement of each single particle; it is the sheer amount of information (collision events) that forces us to resort to a treatment with averages, variances and probabilities. So, whether the outcome is total chaos (apparent or "true"; the observer seldom knows for sure), some semi-deterministic chaotic dynamics (event caught by an attractor, I believe this to be the most common occurrence in nature) or rigid order, we should insist that stochasticity is a matter of description, it is not a fundamental characteristic of the physical world. And this is particularly important when the "gas volume" is a system of living beings, with heredity, senses, signals, social organization, tradition.

Reading Rösner 1990, 1997 leaves me with an impression, that the author at some early stage got overwhelmed by the sheer amount of birds before his eyes, this made him resort to a minimum program of description, ending up in stochasticity. Recoveries and sightings of colour-ringed birds were his best immediate clues to provenience and distribution of birds of known origin, but there was no proper catch outside the arrival and departure months: August and May. Without catches in autumn and winter the Waddensea is a black hole to any field worker; it keeps its illuminating light to itself. Based on his sightings Rösner 1997 notes that birds from more or less limited populations (N. Norway, W. Siberia) do not necessarily stick together on staging or wintering grounds, they disperse in the Waddensea area. But this is not conclusive, it looks the same way everywhere; in the Sound area staging birds and wintering birds mix and compete indiscriminately in foraging flocks, particularly where a wind flat is just being exposed. There comes a time, however, when birds should congregate in order to move on (Fig. 16 and biometry of "Phenology and biometry...", Fig. 2 of "Wintering and spring staging...") or separate in order to economize with food resources (Boere & Smit 1981, Figs. 7, 8 and 9 and text of "Bill length distributions"). What we need to know more about is the mechanism separating birds or bringing them together.

On this point Rösner lacks precision tools that allow him to discern groups of birds in the merged megaflocks alternating between tide flats and highwater resting grounds. The time-tables presented are overall time-tables, the patterns of distribution and movement overall patterns: one single wood of Dunlin where the "trees", the component populations are invisible. It would be unjust to say, that Rösner is totally unaware of these shortcomings; I found the following reflection on "local" difficulties in his introduction: The results from the Waddensea must not be prematurely generalized: several international contacts during the course of this investigation have shown, that individual perspectives bear the stamp of the area of study. Therefore it is necessary to make considerations that go beyond the realms of the Waddensea, maybe even beyond the realms of the flyway. (Rösner 1997). In spite of this, he jumps at conclusion where he should not, this is my main reproach against him.

Rösner's remark should serve as a memento for any study within a restricted area. My own results from the Sound must not be generalized prematurely; I may be in error where I am blind to conditions outside my own context. Still there is one important point to be made, it concerns the expedient chosen by Rösner, and to some extent by Dierschke; it has historical roots. The whole approach resounds of an instructive debate, "Conceptual Issues in Ecology", that took place in the journal Synthese more than twenty years ago (Grene 1980, Levins & Lewontin 1980, Simberloff 1980a, b), with Simberloff advocating a sort of ontological "probabilism" as ecological modernity: biological entities are best treated with statistical techniques, therefore they are stochastic in their general profile. Grene is very harsh on this position; statistics, she remarks, has two sides: it can be used to investigate aleatoric processes, but it also serves to establish probable causal connections. There is no connection between statistical method, unclear causality and ontological indeterminism. It seems to me, this is the mistake of Rösner as well: confronted with the seeming chaos of juvenile movement in the Waddensea he states: their strategy is "trial" - we can leave the errors and strokes of luck aside for a while - the world is stochastic. Viewing the same kind of problem from the vantage point of the Sound area, with seeming navigation of juveniles (or association with adults) at an early age, with staging and encounter, with consistent, symmetrical developments of biometry and morphological characters, I feel inclined to describe the same "chaos" in strict deterministic terms - and extend my views to be valid in the Waddensea area as well.

Some sort of synthesis may be called for in the end, but for the present I think that my data renounce Dierschke's and Rösner's perspective of a "differential migration", followed by juveniles "detecting" favourable areas and surrendering unfavourable ones. Both age-classes have much the same capacities and much the same patterns and move on converging courses, maybe from September and most certainly from October onwards. This convergence of migration and dispersion patterns in turn is of importance when it comes to explaining the diversity in a species that "became divided into several isolated populations during the Pleistocene and has retained a significant amount of intraspecific diversity until the present. (...) Dunlin populations could not become homogenized genetically because gene flow is not extensive enough between them" (Wenink & Baker 1996). These authors argue, that the relative diversity of Dunlin populations is retained as a result of their strong fidelity to breeding-sites, but given the extent of encounter and mixing of populations on wintering-grounds, I doubt that breeding site-tenacity would suffice to keep Dunlin populations separated. By all likelihood the separation between populations is also safeguarded by "fidelity" to an intricate system of staging areas, fattening routines and standard directions for migration relative to a calendar or an internal clock, illustrated e.g. by the very exact arrival (16 November onwards, see "Wintering and spring staging Dunlin...") of wintering adults at the Sound area. There is little room for "trial and error" in this largely deterministic system.

  • To "Studies of migrating Dunlin Calidris alpina in the Sound area, S. Sweden: Introduction"
  • To "Phenology and biometry of Dunlin Calidris alpina migrating by way of the Sound area, S. Sweden"
  • To "Migrating Dunlin Calidris alpina in the Baltic area: the moult issue"
  • To "Risk-prone or risk-averse? Dunlin Calidris alpina migrating with and without moult-gaps in the Baltic area"
  • To "Wintering and spring staging Dunlin Calidris alpina in the south Baltic area"
  • To "Bill-length distributions in Dunlin Calidris alpina"
  • To the "Bill length account"
  • About "adult buff" coverts
  • To the Meissner scale
  • To the literature list A - J
  • To the literature list K - Z
  • Back to start page

    First published 23.2.03, last major addition 17.3.03, last changed 15.12.04, links changed 10.2.07