Moult and migration of the Greenfinch (Carduelis chloris) in SW Scania, Sweden

Results

Fig. 2. Distribution of adult Greenfinches, caught in SW Scania 1979 - 2001. n(female) = 846, median 2 November, n(male) = 609, median 1 November, pentade 61 = 28.10 - 1.11.

Fig. 3. Regression TIME on MOULT SCORE (P + S = 0 - 150) in Greenfinches from SW Scania, material up till and including 17.11. One point may denote more than one bird. Time 1 = 1 August. (Scale: full-grown remige = 10 points; 9 primaries + 6 secondaries; only birds with sheath on P9 or single P9, S6, S5 larger than 0.95 occur as score 150 in the calculations).

Fig. 4. Time distribution of adult Greenfinches with fresh remiges, caught in SW Scania 1979 - 2001. n(female) = 349, n(male) = 304, median of both sexes 2 November; pentade 61 = 28.10 - 1.11.

1. 2c+ female, 26.10.94. S1 - 2 0.95, S3 - 6 old; P8 0.4, P9 old.
2. 2c+ female, 28.10.90. S1 new, S2 0,7, S3 - 6 old; P1 - 6 new, P7 0.9, P8 0.3, P9 old.
3. 2c+ female, 1.11.90. S1 0.95, S2 0.8, S3 - 6 old; P8 0.6, P9 shed.
4. 2c+ female, 3.11.88. S1 - 2 new, S3 0.9, S4 - 6 old; P8 0.95, P9 0.6.
5. 2c+ female, 26.11.91. S2 0.8, S3 - 6 old.
6. 2 2c+ females, 4.12.91. S4 new, S5 - 6 old.

1. 2c+ female, 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.

Fig. 5. Gap, measured in tenths of full remige (inner secondary) length in adult Greenfinches with the last remiges growing, SW Scania 1979 - 2001. n(female) = 212, n(male) = 128.

Fig. 6. Moult score in adult Greenfinches with at least one old remige, SW Scania 1979 - 2001. n(female) = 140, n(male) = 38, average regression line from fig. 3 drawn, double points marked with " etc. Approximative limits, where females with 1, 2 and 3 old remiges are forced to arrest and migrate shown with dashed lines, these limits probably do not surpass TIME = 95 or 100. Right-angled distance to mean regression shown with thick lines. Two birds around 90/90 in the female diagramme had five old remiges.

Fig. 7. The cusp catastrophe as a model for the "extradition" of late moulting Greenfinch females to the migration movement at some latitude in Scandinavia. Numeric values, highly tentative, from fig. 6.

Fig. 8 shows the distribution of adult Greenfinches in at least their 3rd calendar year, caught in Ljunghusen, Lund and Löddesnäs between 2.3 and 20.5 1979 - 2001. The median for males falls on 16 April, for females on 13 April. Note that males are in majority, but if the surplus in early May (30 males) is disregarded the sex ratio becomes 50 : 50 and the median of both sexes falls on 13 April. In April migrating Greenfinches often occur as distinct pairs, the male following a particular female whenever she flies, single males are more common in May, indicating that most of them are local birds with brooding females.



Fig. 8. Distribution per decade (7 = 2 - 11.3) of Greenfinches in at least their 3rd calendar year, ringed in SW Scania 2.3 - 20.5, 1979 - 2001. n(male) = 242, n(female) = 213.

Discussion

I. General I know from the outset what is not allowed: to hold out e.g. the Greenfinch female as "adapted" in her moult, next moving backwards in space and time, destilling a "selective advantage" or pointing out some deft causality against which the moult process rubs itself, like a cat against its master's leg. Even the most conscious and conscientious Darwinists are unable to resist this basic, empty gesture. Second: I find the notion of "adaptive programming" in e.g. the moulting bird of minor interest; the internal readiness is indisputable, but it leads the thinking astray: the transition to moult or migration is at any rate not mechanically, automatically instructed from the level of the genome. Third: I cannot avoid adaptive figures altogether, they sneak in underhand, a system "adapts" itself through its dynamics. I can insist, however, that this adaptation is not rooted in a fix code, but in the capability of the systems dynamics to react on and promptly handle signals from the inside and the outside alike. A program-controlled system is not always structurally stable (producing a certain result also when perturbed), but an open, dynamical, information-processing system may be constructed for structural stability. This is the crucial point, the one that breaks the "logic" of Darwinism. He or she who is uncertain about the content of this objection should keep clear of Darwinism and avoid terms like "adaptive" or "selective" (they need to be purified first).
In order to emphasize that I move against the current, that my direction of thinking is opposite to the Darwinist direction, I choose to call the female Greenfinches "misfitted" in their moult, consequently, systematically misfitted. An excellent term, provocative in its double connotation (it also announces that the bird will not be pulled through the current adaptive threshing machine). This is my starting-point: the mis-adapted moult of the Greenfinch female.

II. Role of the catastrophes There are "catastrophic" mechanisms in bird populations (but in all other biological systems as well - down to the hibernation of a collapsed system as "spores"), dynamical courses that interrupt a breeding failure, orchestrate a second attempt if there is time, expel moulting birds if their time is running out. A sort of "emergency brakes" or abortions (spontaneous abortion contains all that is typical of catastrophes), that gather up the crowds when something has gone wrong, when holocaust is an option if worst comes to the worst. This is a systemic quality, evident from the fact, that the interruption takes place synchronously, imperatively, in the way gravel is tipped from a lorry. The catastrophe is the ultimate self-preservation of the population; insofar it is a foundation bolt. These - in the population, in the dynamics of the system rooted - catastrophes are "creative", they create the more or less permanent time limits of the system and the synchronization between different states/phases of the annual cycle. We may for the present avoid the question where exactly they are housed, if they are mainly "external" or mainly "internal" - but it is fairly obvious that they must be both; their foundation is the interaction between rather primitively coded (linear) information from the inside and the outside alike.

III. Catastrophes and general dynamics In Sweden the Greenfinch has increased its numbers and expanded its distribution northwards, this development has been particularly pronounced in the second half of the century (SOF. 1990). There are many similar "episodes" in the zoogeographical history of Sweden in the 20th century: the expansions of Black-headed Gull (Larus ridibundus), Collared Dove (Streptopelia decaocto) and Scarlet Grosbeak (Carpodacus erythrinus) for three examples. The pioneers of such expansions are always extremely "misfitted" to start with, they are put out in the cold in more than one respect. In all contexts of this kind catastrophes manifest themselves: interrupting breeding that has suffered from low temperatures, rain, drought, or lack of food, expelling Long-tailed Skuas (Stercorarius longicaudus) to the oceans, sending Slender-billed Nutcrackers (Nucifraga caryocatactes macrorhynchos) across continents in the middle of the summer - and expelling late moulters before their chances of getting anywhere are completely spoiled. (The invasion is a well-known manifestation of catastrophic expulsion; its phenomenology has never been quite penetrated - here the moult occupies a central role - nor has it been connected with other types of interruptions, sharing the same dynamical background). The interrupting catastrophe is preserving, it salvages the remnants and lays out the fundament for another attempt, be it with weakened numbers. Or expels a remnant with weakened but not entirely unfit condition. In this function it is an important agent in subarctic, but also in temperate areas, it keeps a population from fighting too hard against the odds. (In arctic areas a species should fight, with extra fat, fur and feathering, but catastrophes are beneath the surface even here, only the thresholds are higher).
Thus, the most important function of a catastrophe is interruption. But there may be secondary effects as well: it is capable of moving time-limits, insofar as there are internal clocks, running from the interruption of last year's breeding or moult (references here). In this way the catastrophe may turn out to be creative, or, using a term borrowed from Levontin (1978): problem-solving (not applied to catastrophes by Levontin). Progressively creative, in a teleonomic, vaguely goal-seeking process.

An expansion phase in a bird species is a period of intense dynamics, and catastrophes are at work on several levels: moving spring migration backwards in the calendar, synchronizing a tricky breeding, and piloting an adequate number of adults from moult to consecutive phases of the annual cycle. During such periods the state equations of the system lead to a particular zone with numerous catastrophes. Later on the population may show a more static behaviour, the dynamics loses its potency, becomes in a way powerless, and numbers often decline at the same time. Now the state equations lead to another, non-catastrophic zone, this is what may have happened to the earlier mentioned species Black-headed Gull, Collared Dove and Scarlet Grosbeak. One hypothesis immediately offers itself: populations (or perhaps rather: central population parameters) pulsate a little in this respect, and the strong, expansive populations are characterized by being located to the "ultradynamic" zone, where several broad, displacing/interrupting catastrophes are active and "cut" in the different phases of the annual cycle. Here is a regulatory mechanism, that sees to it that the informational base, the starting-points rendered by the genome, need not be modified at once (or at all), perhaps prematurely. (And how a basic change is effected, if it comes, is an open question). As a consequence a higher frequency of arrested moult could be expected in an expansive, dynamic population, more of strictly regulated (suspended or complete) moult in a static or receding. This could be worded even more incisively: The strong population is more "misfitted", its material is richer and it "plays" more with this material, sacrifices more of either sex; the static or receding population is more "adapted" and anxious, also about its material.

IV. Greenfinches: general The annual counts from 27 years at Falsterbo (Roos 2000) indicate that the Greenfinch populations migrating by way of S. Sweden hold their position as late as 1999, there are no signs of a "Black-headed Gull decline". Second, table I (above) indicates a stable, surprisingly high frequency of late or arrested moult. And the migration diagrams (Figs. 2, 8) show how all phases have been stretched and displaced; spring migration starts very early and ends late in the season - be it late in April rather than in May. The autumn migration of adults doesn't culminate until one or a few days into November and lasts throughout the month. Historical material (Rudebeck 1950, Ulfstrand et al. 1974) indicates that there have been changes on the part of SW Scania, while the amputated diagrams in Edelstam (1972) suggest large-scale November migration at Ottenby as early as the 1940's. I do not know which source and which supposition to trust and choose here.

As early as the 1940's day length or photoperiod were pointed out as ultimate factors triggering moult, and Greenfinches were involved in the first attempts to manipulate the annual cycle by means of artificial light (Damsté 1947). There followed a long development, prompted above all by the Germans Berthold & Gwinner, with a parallel development in America and intense exchange between the two camps. Berthold and Gwinner were trying to establish a hereditary "circannual periodics", running autonomously, exceeding or falling behind the calendar year, and reacting to external stimuli. The role of the photoperiod and the "internal clock" have been established for decades now, but what has often escaped notice during this long development is the fact, that much research took place under laboratory conditions, and that the "field contact" of Berthold and Gwinner (e.g. Gwinner 1968, Gwinner, Berthold & Klein 1972, Berthold, Gwinner & Klein 1970, 1972a, b, Berthold 1974, 1975, Berthold & Schwabl-Brenzinger 1982) left something to be desired; there are several examples where they stretched reality to make it fit into their models (for one example see Högstedt & Persson 1982). I still experience this lack of field contact and tendency towards mechanical thinking in many modern writings about circannual periodics, "internal clocks"; whenever I scrape on a species (any species) I find another moult schedule, another "stretching" and another timing than suggested in handbooks and papers of even recent date. Six years of Dunlin study in the Sound area were enough to shatter the petrified image of Dunlin moult (Greenwood 1983, Holmgren et al. 1993, to: Moult of Dunlin Calidris alpina in the Sound area, S. Sweden); it is hard to avoid the feeling, that everything remains to be done in this field. And the thing that happens again and again is characteristic: a rather square picture is dissolved; conditions turn out to be far more flexible - not an indiscriminate mix, but regulated and flexible - than originally assumed. An expression like "adaptive temporal programming of moult and migration disposition" (Berthold & Schwabl-Brenzinger, 1982), very characteristic of the whole sphere, steers thought and eye in the wrong direction; the researcher tends to adapt to his presumed adaptation and filter away or be blinded to the perhaps most interesting material. In all these cases I can feel the ghost of Darwinism in the background: it only sees individuals, racing between stops for individual optimization, but is blind to their resting (maybe not always safely) in the cradle of the system/population, and having their essential movement through this cradle.

V. Greenfinches: specific I have seen 2y, non-breeding Greenfinch males with shed first primary in the last days of June (but there are no moult-cards). In Scania the average moult start falls on or shortly after 1 August, and among birds migrating by way of Scania in October/November a few have not started moulting until 1 - 15 September (and there must be such late birds of Scanian origin, too). Newton (1968) gives 5 August as mean starting date for 115 investigated British Greenfinches, the 95 %-interval falling between 25.6 and 14.9, i.e. two and a half months! The span should be much the same in Scandinavia (the summer material in this paper is extremely poor), this in turn is packed together into an effective adult migration period (where the most distant migrants have already flown 1400 kms) of little more than a month. Furthermore Newton (1972) estimates that all secondaries grow for 20 days, the same valid for the short primaries (P1 - 4), while the growth of P9 lasts 30 days. If all this is to be bundled into a 85-day-period, remiges must grow with appr. 50 % overlap in the following way: P1 - 3 20 days; P4 - 6 overlapping with these and with S1 - 3 +10 days; S1 - 3 +10 days, P7 + S4 overlapping with S1 - 3 and the last primaries +13 days; P8 - 9 plus S5 - 6 +30 days, summing up to 84 days. According to Newton (1968) on average 2.6 primaries grow in Greenfinches from the shedding of the third until the seventh is full-grown. If this course of events is to be accelerated even more (after, say, moult start on 15 September) the bird has to grow more remiges at the same time, and risk to be handicapped in flight, or arrest prematurely. My first and most important hypothesis goes: this interruption is effected through catastrophic dynamics. The presence of catastrophes is not "proven" by the material presented here - there is no point in calculating with four birds having four old remiges, I could need 5.000 moult-cards - but I think that the presence of catastrophes has for the first time been made probable¹. The magnitude of this probability is a decision for the individual reader to make; I think I should not argue more in this particular case, but continue to other species, other contexts, and submit new material, illustrating the same phenomenon from other angles.
1. One problem is, that the interrupting catastrophes in, say, Norrköping and Uleåborg, are most certainly displaced relative to each other; hence the drawn-out bands of data points among migrating birds in Scania. But again it is obvious how, the same way as in the case of Yellow Wagtails, there is a sort of all-embracing harmonization at some latitude; practically all birds suddenly have the same readiness for migration, the same flight capacity, the same fattening and move together in the same main direction. The complexity of this synchronization makes the head go round! Of course calculations should preferrably be based on "pure" materials from Finland, N. Sweden, Norway, central Sweden etc., but recording moult is a difficult thing when the whole population is ready for departure. It is practically always necessary to "beat windward" when researching moult in the field, making the best of the existing conditions.

From a temporal point of view Greenfinch males are without exception in a better position than females; when migration starts they are in majority in pentades 58 and 59 (Fig. 2), before females - in overall majority during migration - take over. Two thirds of the males have had completely fresh remiges, while a majority of females still has growing remiges on migration in SW Scania. The average gap of all moulting categories was 0.4 - 0.5 remiges (secondaries) with standard deviation the same order of magnitude. When the median date of migration approaches, all moulting categories of the diagram TIME on MOULT SCORE are packed together and the regression becomes practically vertical, here conditions remind of conditions in Yellow Wagtail moult - to: Moult of Yellow Wagtails (Motacilla flava) in Sweden. But the remiges of Yellow Wagtails grow faster in the final phase, while Greenfinches lose momentum, which makes their moulting pattern reminiscent of the pattern in e.g. Waxwing (Bombycilla garrulus), Siskin (Carduelis spinus) and Crossbills (Loxia sp.) in "invasion years". This is an important observation, I regard it as a confirmation that the rather rigid temporal framework in "regular", long-distance migrants was established, and is managed by catastrophes of the kind that "send out" an invasion at any time of the year. The Greenfinch has much to contribute in the research of such time-patterns, and it is easy to catch, maybe the species standing closest to European bird-ringers.

SUMMARY:

Moult and migration in Greenfinches (Carduelis chloris) in SW Scania, Sweden. A material comprising 1171 moult-cards from SW Scania, mainly from late autumn (overall time distribution: Fig. 2) is analyzed. The summer material is poor, and the main focus is directed on the shift from the moult state to the migration state. Following Newton (1968, 1972) and Ginn & Melville (1983) a mean start by August 1st and a mean duration of (at least) 85 days for remige moult is assumed also in Scandinavian Greenfinches. These values may vary from south to north; the necessary material to judge in this matter is not at hand. Two thirds of all males had fresh remiges, while a majority of the migrating females still had at least one growing remige. In birds with growing remiges the mean gap was 0.4 - 0.5 inner secondaries, standard deviations the same order of magnitude. This is probably what any migrant can sustain; in extreme cases the gap amounted to as much as 1.5 inner secondaries. Late moulters (in particular females; see Fig. 6) arrested and left 1 - 4 inner secondaries unmoulted. There is no case of unmoulted primaries from early November onwards. When Greenfinches enter the migration state, the growth of feathers seems to come to a halt, at least the growth rate is extremely low. S6 grown to 0.95 - 0.98 of full length is often met with in adult males in spring, and a female arresting after S3 had this remige only 0.9 of full length on 21 August in the following summer. It is suggested, based on Fig. 6, and with reference to Moult of Yellow Wagtails (Motacilla flava) in Sweden that the moult process is to a high extent regulated, governed by information, and that the necessary interruption ("extradition" of late moulters) is taken care of by catastrophic dynamics Elementary catastrophe theory, an introduction, always potentially present below the surface in any group of living beings, linked together by communication. It is also suggested, that these catastrophes originally were instrumental - and continuously are instrumental - in creating the temporal pattern of e.g. an annual cycle in any living organism.

Litteratur/Literature:

BERTHOLD, P. (1974): Circannuale Periodik bei Grasmücken (Sylvia). III. Periodik der Mauser, der Nachtunruhe und des Körpergewichts bei mediterranen Arten mit unterschiedlichem Zugverhalten. J.Orn. 115: 251-271. º BERTHOLD, P. & I. SCHWABL-BRENZINGER (1982): Adaptive temporal programming of molt and migratory disposition in two closely related long-distance migrants, the Pied Flycatcher (Ficedula hypoleuca) and the Collared Flycatcher (Ficedula albicollis). i: PAPI, F. & H. G. WALLRAFF (eds.): Avian Navigation. Springer Verlag, Berlin, Heidelberg. º DAMSTÉ, P. H. (1947): Experimental modification of the sexual cycle of the Greenfinch. J. Exp. Biol. 24: 20-35. º GINN, H. B. & D. S. MELVILLE (1983): Moult in Birds. BTO Guide 19. º GWINNER, E. (1968): Circannuale Periodik als Grundlage des jahreszeitlichen Funktionswandels bei Zugvögeln. J. Orn. 109: 70-95. º GWINNER, E. (1969): Untersuchungen zur Jahresperiodik von Laubsängern. Die Entwicklung des Gefieders, des Gewichts und der Zugunruhe bei Jungvögeln der Arten Phylloscopus bonelli, P. sibilatrix, Ph. trochilus und Ph. collybita. J.Orn. 110: 1-21. º GWINNER, E., P. BERTHOLD & H. KLEIN (1972): Untersuchungen zur Jahresperiodik von Laubsängern. III. J. Orn. 113: 1-8. º HASSELQUIST, D. et al. (1988): The seasonally divided feather moult in the Barred Warbler Sylvia nisoria - a new moult pattern for European passerines. Orn. Scand. 19: 280-286. º HAUKIOJA, E. (1971): Flightlessness in moulting passerines in Northern Europe. Orn. Fenn. 40: 101-116. º HEDENSTRÖM, A. & J. PETTERSSON (1987): Migration routes and wintering areas of Willow Warblers Phylloscopus trochilus (L.) ringed in Fennoscandia. Orn. Fenn. 64: 137-143. º HOLMBERG, T. (1992) The primary moult of the Brambling Fringilla montifringilla evaluated by four different methods. Orn. Sv. 2: 139-156. º HÖGSTEDT, G. & C. PERSSON (1982): Do Willow Warblers Phylloscopus trochilus of northern origin start their autumn migration at an earlier age than their southern conspecifics? Hol. Ecology 5: 76-80. º LEHIKOINEN, E. & P. NIEMELÄ (1977): (Moult study on passerines). Lintumies 12: 33-44. º LEVONTIN, R. (1978): Scient. Am. Sept. 1978. º LUNDBERG, P. & L.-O. ERIKSSON (1984): Postjuvenile moult in two northern Scandinavian Starling Sturnus vulgaris populations - evidence for difference in the circannual time-program. Orn. Scand. 15: 105-109. º NEWTON, I. (1968). The moulting season of some finches and buntings. Bird Study 15: 84-92. º NEWTON, I. (1972): Finches. Collins, London. º OTTOSSON, U. & F. HAAS (1991): Primary moult of the Brambling Fringilla montifringilla in northern Sweden. Orn. Sv. 1: 113-118. º PERSSON, C. (1977): The early stages of the postnuptial moult in the White Wagtail Motacilla alba. Orn. Scand. 8: 97-99 º PERSSON, C. (1983): Kring en grönfinkpassage hösten 1982. Fågelstudier 1: 75-81. º ROOS, G. (2000): Sträckfågelräkningar vid Falsterbo hösten 1999. Anser, suppl. 44: 65-91. º SOF (1990): Sveriges fåglar. 2:a uppl. Stockholm. º STRESEMANN, E. & V. (1966): Die Mauser der Vögel. J. Orn., Sonderheft.

The first version of this paper is written by Christer Persson and published on the web on 6.12.01, English version 28.12.01. An important part of the material comes from ringings done by Ulf Lundwall in Lund and Löddesnäs, interesting moult-cards from two occasions by Peter Olsson at Kämpinge in the autumn of 1994. I need not add: I would be glad to see and calculate more material, preferrably thousands of moult-cards from the whole of Europe! I am interested in the number of old remiges in spring, I am interested in weights of moulting birds (but do not weigh where they are fed). Use decimal scale when recording moult, this gives better resolution, important in the study of details from the final phase! Summer material from the whole of Scandinavia, allowing calculations of the average moult-start, would be particularly welcomed.

APPENDIX I: Recruitment area of moulting Greenfinches (Carduelis chloris) in this study.

Till artiklar om ruggning/To articles on moult

Till engelska innehållsförteckningen/To English contents