Introduction: A long-term study of Yellow Wagtails (Motacilla flava) at roosts in SW Scania, Sweden

For readers from other continents: The Yellow Wagtail Motacilla flava breeds with numerous subspecies throughout the Palearctic. The subspecies flava breeds in the southern half of Sweden, in southern Finland and in southernmost Norway, the subspecies thunbergi in northern Sweden, northern Finland and most of Norway, in addition there are hybrids in a zone of encounter in all three countries. Adults moult completely in late summer and the end of moult is exactly synchronized with the start of migration.

Background

The Yellow Wagtail (Motacilla flava) is a numerous migrant in South Sweden in autumn, as noted already by Rudebeck (1950) in his pioneering work from Falsterbo. When discussing annual counts of migrants at Falsterbo, Ulfstrand (1957) remarked that this species "seemed to be the most regular of all", the general stability applying to phenology as well as overall numbers. For certain, strict circannual rhythm and strict migration schedule are traits common to all long-distance migrants, but the stability of numbers gives the Yellow Wagtail a top position even in this context. At least two aspects of this stability are open to field study with relatively simple means:

The pronounced synchronization of the consecutive events breeding, moult and migration. Synchronization is a common theme to practically all articles dealing with breeding, moult and migration in "Fågelstudier"; the different sections will gradually be summarized and catastrophe theory applied to them.
The absence of population "crashes" connected with breeding failure, migration catastrophes or severe conditions in the wintering area. Population density is a parameter in all catastrophic dynamics at the population level, and it is reasonable to assume some sort of connection between strict synchronization and stability. The relationship is dialectic, there is an influence in both directions: stability contributes to synchronization, but the really interesting question in the present context is the influence in the opposite direction: how is stability (of population levels) promoted by synchronization?

It has been suggested that a very exact "endogenous clock" (Gwinner 1968, Berthold, Gwinner & Klein 1972a, b, Berthold 1974) lies behind the timing of e.g. moult and migration, and that it is specific to populations, but also must involve coordination between populations (cf. the hypothetical "internal timer" of Curry-Lindahl 1958, 1963, the very interesting questioning of this (omitted in most references; ornithologists can not cope with two hypotheses simultaneously!) by Owen 1969, also see Ward 1964, Moreau 1972). There must be some essence of physical truth here, but it is wrong to visualize each bird running like a Wonderland rabbit with a clock before its eyes; when it comes to group behaviour the clock construct is fictional and subordinate and does not capture the true and really interesting relationships.

The synchronism and coordination on the autumn side of the annual cycle have been main topics of interest in a long-term study (1964-99), where large numbers of migrating Yellow Wagtails were caught, ringed and investigated for moult at roosts in SW Scania. In addition the same work has provided information on breeding and wintering areas as well as on migration routes of the populations involved; other conditions underlying the general stability can be assessed against this background. So, this paper deals with interaction, between physiological processes, between populations, between different segments of the annual cycle, and the overall issue is the relative stability resulting from this interaction (and synchronous action) at different levels.

Material and methods

Migrating Yellow Wagtails congregate in large roosts in autumn. In SW Scania, north and east of Falsterbo peninsula, such roosts are particularly common; during peak migration the whole coastal area from 55º 24' (Falsterbo) to at least 55º 53' N (Landskrona) will shelter hundreds of thousands of birds overnight. In this area, particularly at Foteviken (c55º 27' N, 13º 00' E) and Löddesnäs (55º 44' N, 13º 00' E), Yellow Wagtails were ringed on a large scale between 1964 and 1999. At both sites the birds were caught with six to fifteen 9-meter, 4-shelf mist nets, the number of nets noted on each catching occasion. The birds were ringed and investigated directly from the nets by standing ringers, after sunset with forehead torches as light-source. Under normal weather conditions age, sex and subspecies (mainly in males) were noted, if catches were very large or bad weather seemed to be ahead, a portion of the birds could be ringed without any determination and let free as soon as possible in order to lessen the strain on nets and birds caught in them.

Between the Foteviken area and Klagshamn (c55º 31', 12º 56'), there are about 25 separate reeds that will serve as roosts for Yellow Wagtails. (Fig. 1) The actual choice - with new point of gravity from night to night - is decided mainly by wind-direction, the descending flocks always approaching and entering the reed-beds against the wind. The presence of predators will affect the choice to some extent as well. By the end of August as much as 10.000 Yellow Wagtails will spend the night in the overall area. Given the shifting preferences, a tape lure is very helpful if the ringer wants to catch as many birds as possible, in addition the birds will be caught earlier in this way, and daylight always facilitates both handling and race determination. Still, a tape lure was not consistently used at Foteviken until 1986, so the material collected here before that year is "undisturbed" by this remedy. In contrast, tape lures were used at Löddesnäs from the very first year, and this is reflected in catching numbers; they are close to the optimum of the ringer, reduced only by periods of bad weather. The setback with lures is that catching numbers will not reflect the number of birds spending the night in an area; if the nets are properly set the catch will at best reflect the number of nets used. With a lure it is as easy to catch 300 birds out of a thousand as it is to catch them out of ten thousand (or even easier).

ringingsites

Fig. 1. Location of major roosts, where Yellow Wagtails have been ringed 1964-99. Many birds leave Sweden by way of "Nabben", the SW cape of the Falsterbo peninsula.

In the Foteviken area 53.100 Yellow Wagtails were caught between 1964 and 1999; 1.582 not properly aged between 1964 and 1966, from 1967 onwards 1.060 Motacilla flava flava in nest plumage, 41.744 "plain" juveniles, mostly in 1st winter plumage, 7.426 adults and 1.288 unaged birds. Christer Persson (with many helpers) started this catch in 1964, Peter Olsson and Kerstin Norrman joined from 1986-87 and made independent contributions in the early nineties, they in turn assisted by Per Nothagen; in the late nineties the catching effort gradually faded away. By then Yellow Wagtails had been caught on 714 separate occasions, some hundred of these including only a couple of birds.

Parallel to this, Ulf Lundwall started ringing Yellow Wagtails at the mouth of River Lödde (Löddesnäs, see Fig. 1) in 1979; up till and including 1997 (aided by Håkan Skoglund, in addition Peter Olsson ringed here on five occasions in 1995) this resulted in 25.978 birds ringed on 255 occasions: 21.241 juveniles, 4.731 adults and 6 unaged. Unlike Foteviken/Klagshamn Löddesnäs is one restricted reed, serving as roost for many species, including Starlings (Sturnus vulgaris) and hirundines, from this point of view it is the better site; single catches of 300 birds are easily obtained here with little more than ten nets. It should be added, that in many years activity at Löddesnäs was restricted to weekends (Friday - Sunday), while Foteviken ringers have had more freedom to select the days or periods of peak migration.

All in all 79.078 Yellow Wagtails were ringed in 36 years at these two sites (i.e. an average of more than 2.000 per year over a period of three and a half decades), resulting in 173 long-distance recoveries (0.22 %) and some 500 local retraps within and between roosts. In addition 9 birds with foreign rings were controlled, and 17 birds ringed elsewhere in Sweden. Furthermore 506 moult-cards related to actual postnuptial moult (from first remige shed to sheath on ninth primary and inner secondaries) were collected at Foteviken, Klagshamn and Löddesnäs. These combined data: distributions of race- and age-determined birds, recoveries and moult cards form the material of this paper.

Results

1. Distribution of catch, ages and subspecies

The catching effort

Hunting traditionally starts on September 1st in Foteviken; the general anarchy and turmoil created by hunters has made us avoid the roosts on this day and often on the next day as well. On 1 and 2 September combined there are c15 catching occasions less than expected, this absence affects the median date by almost a full day. Löddesnäs is a nature reserve and exempt from hunting. Overall catching numbers for the Foteviken area are given in Fig. 2, for Löddesnäs in Fig. 3, the number of catching occasions at both sites in Fig. 4. The overall median date for the Foteviken area is 28 August (1y 28 August, 2y+ 30 August), for Löddesnäs 29 August (both 1y and 2y+ the same date).

Fotenumbers

Fig. 2. Overall catching numbers per pentade (pentade 49 = 29.8 - 2.9) of 1y, 2y+ and 1y+ Yellow Wagtails in the Foteviken area, 1967 - 1999; n = 51.518. Median date shown by arrow.

Löddenumbers

Fig. 3. Overall catching numbers per pentade (pentade 49 = 29.8 - 2.9) of 1y and 2y+ Yellow Wagtails at Löddesnäs, 1979 - 1997; n = 25.978. Median date shown by arrow.

no.occasions

Fig. 4. Number of catching occasions per pentade at Foteviken and Löddesnäs.

In Table I. more detailed information about the catch is given for each single year and each site.

Year, site	Total	1y	2y+	1y+	Main interval	No. of catches in main interval	Average catch/occasion in main interval	Average no. of nets (no. of occasions)
FOV, 1967	1035	895	140	69	22.8-12.9	13	75	7.0 (7)
FOV, 1968	904	388	37	479	20.8-6.9	10	82	-
FOV, 1969	1138	803	130	205	19.8-12.9	12	87	-
FOV, 1970	1390	1241	144	5	17.8-11.9	18	72	-
FOV, 1971	843	722	118	3	19.8-13.9	17	46	-
FOV, 1972	665	578	52	35	28.8-11.9	12	53	-
FOV, 1973	1723	1530	192	1	15.8-8.9	23	74	-
FOV, 1974	1314	1063	229	22	19.8-10.9	17	76	-
FOV, 1975	2200	1881	305	20	14.8-9.9	20	104	6.8
FOV, 1976	2437	2242	176	19	14.8-12.9	24	97	6.2
FOV, 1977	1913	1548	336	29	15.8-11.9	23	82	7.3
FOV, 1978	925	823	94	8	16.8-8.9	10	90	7.6
LNÄ¹ 1979 FOV 1979	428	369	59	-	25.8-4.9	4	107	6.5
LNÄ¹ 1979 FOV 1979	902	789	110	3	17.8-9.9	15	58	7.3
LNÄ 1980 FOV 1980	482	452	30	-	1.9-10.9	5	97	7.4
LNÄ 1980 FOV 1980	927	800	117	4	25.8-11.9	16	58	7.5
LNÄ 1981 FOV 1981	1157	907	250	-	7.8-16.9	11	104	(11.3)²
LNÄ 1981 FOV 1981	1066	830	222	14	19.8-9.9	15	69	8.9
LNÄ 1982 FOV 1982	937	820	117	-	15.8-5.9	7	133	9.0
LNÄ 1982 FOV 1982	1580	1154	394	32	16.8-10.9	18	86	10.8
LNÄ 1983 FOV 1983	2033	1737	296	-	17.8-14.9	13	152	8.3
LNÄ 1983 FOV 1983	1892	1575	287	30	15.8-14.9	22	84	11.2
LNÄ 1984 FOV 1984	1625	1286	338	1	9.8-10.9	13	124	5.5
LNÄ 1984 FOV 1984	1471	1062	341	68	20.8-10.9	14	104	12.6
LNÄ 1985 FOV 1985	1861	1458	403	-	15.8-2.9	15	97	6.1
LNÄ 1985 FOV 1985	1099	885	208	6	19.8-10.9	17	64	8.9
LNÄ 1986 FOV 1986	2281	1994	287	-	11.8-13.9	18	126	6.7
LNÄ 1986 FOV 1986	3775	3110	630	35	13.8-13.9	30	126	9.3
LNÄ 1987 FOV 1987	2750	2098	648	4	16.8-8.9	17	161	9.1
LNÄ 1987 FOV 1987	2658	2094	546	18	16.8-13.9	23	114	10.3
LNÄ 1988 FOV 1988	2837	2321	516	-	14.8-17.9	21	135	9.2
LNÄ 1988 FOV 1988	4914	4147	750	17	9.8-17.9	33	147	10.3
LNÄ 1989 FOV 1989	1417	1159	258	-	13.8-14.9	13	108	7.5
LNÄ 1989 FOV 1989	4504	3851	615	38	10.8-14.9	31	141	12.0
LNÄ 1990 FOV/KLH 1990 LJH 1990	1063	911	152	-	13.8-5.9	8	132	6.8
	2641	2417	213	11	10.8-10.9	27	97	-³
	898	823	64	11	17.8-12.9	20	45	4.9
LNÄ 1991 FOV/KLH 1991	1030	848	182	-	20.8-9.9	8	123	8.0
LNÄ 1991 FOV/KLH 1991	880	764	108	8	26.8-13.9	12	73	7.9
LNÄ 1992 FOV 1992	872	669	203	-	18.8-10.9	8	107	5.4
LNÄ 1992 FOV 1992	1140	957	174	9	14.8-9.9	10	111	8.0
LNÄ 1993 FOV/KLH 1993	637	544	93	-	22.8-12.9	7	89	5.7
LNÄ 1993 FOV/KLH 1993	1455	1158	219	78	15.8-16.9	13	107	6.1
LNÄ 1994 FOV 1994	1642	1332	310	-	17.8-7.9	11	146	6.3
LNÄ 1994 FOV 1994	1277	1054	211	12	10.8-17.9	15	82	7.6
LNÄ 1995 FOV 1995	497	368	128	1	17.8-19.9	6	83	-
LNÄ 1995 FOV 1995	738	651	81	6	16.8-12.9	9	81	7.1
LNÄ 1996	971	797	174	-	25.8-14.9	7	136	6.0
LNÄ 1997 FOV 1997	1458	1171	287	-	13.8-3.9	11	131	7.3
LNÄ 1997 FOV 1997	593	521	67	5	21.8-31.8	7	85	6.7
FOV 1999	427	376	49	2	20.8-10.9	6	69	6.2

Table I. Catching effort at Foteviken (FOV), Löddesnäs (LNÄ), Klagshamn (KLH) and Ljunghusen (LJH) 1967 - 99. Note 1: 3 out of 4 catches at Saxtorp, a few km N of Löddesnäs. Note 2: 2 - 3 extra nets set for other species included; the value 8 nets used in calculations. Note 3: No exact information on nets, but seldom more than eight used.

Age distribution

In S. Sweden most juvenile Yellow Wagtails visiting roosts by the end of July are still moulting body-feathers. When the body moult is finished, juvenile greater coverts are still retained by all juveniles of thunbergi and hybrid thunbergi/flava type, and by most juveniles of flava type; the exception being a few % of the local juveniles ("southern flava", see below) and single juveniles of "standard flava" type, that will moult all greater coverts as well. The juvenile covert has a white or buffish-white edge, clearly demarcated from the core of the feather, while adults have buff, olive-buff or yellowish-buff edges gradually fading into the centre. This is a reliable and easily discernable age criterion, working even under poor light conditions, and whenever there has been suspicion of a completely moulted juvenile (edges of greater coverts fading into the centre, but tending to be more whitish or yellowish than in adults) and difficulties with other criteria, like coloration of iris and inside of bill, it has been denoted as "1y+" by the Foteviken ringers. There are few 1y+ from Löddesnäs, and possibly some small error in age determination here. Fig. 5 presents age ratios and standard deviations from pentades 45 - 52 in Foteviken and Löddesnäs; the overall adult ratio at Foteviken (n = 51.518) has been 14.8 ± 0.16 %, at Löddesnäs (n = 25.972) 18.2 ± 0.24 %, the rate of Löddesnäs highly significantly larger (Z = 11.89, p<0.001) than the rate at Foteviken.

age ratios

Fig. 5. Adult ratios ± 1 s.e. per pentade at Foteviken and Löddesnäs. The differences between mean rates of pentades 45 - 48, 50 and 52 highly significant (Z = 2.79, 8.48, 3,85, 8.73, 5.40 and 3.52 resp.; p<0.01), the differences of intervals 49 and 51 almost significant at the 5 % level (Z = 1.77 and 1.93 resp.).

Pentades 47 - 51, i.e. the period 19 August to 12 September, see the main part of adult migration, with the adult share reaching or exceeding 15 %. The distribution of adults is very regular (also see Figs. 16 - 23); if there is some skewness to the left in the overall distribution (1y + 2y+), it is connected exclusively with the distribution of 1y birds.

From Fig. 5 it is clear, that any calculated adult ratio will depend on the period when the underlying data were collected. But even with good planning and forethought sampling errors can not be avoided; if juvenile catches in pentade 46 are successful because of good weather, and adult catches in pentade 49 unsuccessful because of bad weather, the overall ratio will be distorted. This has happened to us more than one year, when strong westerly winds frustrated all plans for the second half of the catching season. So, any calculated age ratio should be examined and cross-examined before being admitted and used for further purposes. With this reservation age ratios from Foteviken 1967 - 99 and Löddesnäs 1979 - 97 are presented in Fig. 6; most of the values are based on sound materials and are highly relevant - a few are misleading and will be pointed out later on.

Fig. 6. Adult ratio (± 1 s.e.) of the total catch at Foteviken 1967 - 1999 (left) and Löddesnäs 1979-97 (right).

In spite of possible defects, three features deserve to be brought out in the above diagrams:

Significant jumps (order of magnitude 1 - 5 standard errors) in the age ratio are a recurrent feature in the Scandinavian Yellow Wagtail population.
The general impression given by both diagrams is, that strong year classes put their mark on the population development for one or several consecutive years; after a year with good reproduction there will often follow a year or two of consolidation, before another strong year class is fostered.
Between 1983 and 1992 the tendencies of both diagrams are much the same.

Still, the diagrams are of limited value in the above, "raw" state. At Löddesnäs, there were no catches in pentade 48 in 1982, and only one in pentades 49 and 50 of that year. The low adult ratio of this site in 1982 is an artefact, so is the high ratio in 1995, when there were no catches in pentades 47 - 49. Hence, the catching numbers of different years and pentades must be evaluated and - where possible - corrected before they can serve as e.g. an index of population development. This correction must proceed along several roads; the procedure is illustrated in Figs. 7 - 11, depicting the correction of values from the years 1982, 84, 86, 92, 93 and 95. Löddesnäs 1980 and 1982, Foteviken 1996 and 1997 could not be corrected; the samples are too few and too biased. Finally, values from both sites are brought together in Fig. 12; this diagram includes uncorrected values from years with very good sampling as well as corrected values from years with deficient sampling.

age ratios 1982

Fig. 7. Adult ratios per pentade at Foteviken (circles) and Löddesnäs (squares) in 1982. In this year many adults were belated with an outbreak of 126 2y+, 176 1y (42 % 2y+) in Foteviken on 7 and 8 September, while the last catch at Löddesnäs took place on 5 September, in addition there was only one catch in pentade 49. The value from Foteviken is based on 18 catches in pentades 46 - 51, the regression shown is highly significant (p<0.01 in constant and regr. coeff.), the value from Löddesnäs is faulty, and correction will add nothing.

age ratios 1984

Fig. 8. Adult ratios per pentade at Foteviken (circles) and Löddesnäs (squares) in 1984. Löddesnäs missed an outbreak of 2y+ birds with 93 2y+, 127 1y (42 % 2y+) in Foteviken on 5 September, no catch at Löddesnäs between 4 and 11 September, good coverage in other pentades. The values from both sites have been regressed according to the figure and the percentage for pentade 50 at Löddesnäs has been changed to 34.9 (regression; both constant and regr. coeff. highly significant). In addition mean values 7.0 (Foteviken) and 12.3 % (Löddesnäs) have been used for pentade 46, with no catches at either site. Finally the percentages, true and assessed, have been applied to the average distribution of 1000 birds from pentades 46 - 51 (95 % of the annual catch) at Foteviken (97 - 163 - 240 - 195 - 199 - 106) and Löddesnäs (92 - 194 - 203 - 238 - 178 - 95). The corrected values are 21.7 % (Foteviken) and 24.3 % (Löddesnäs).

age ratios 1986

Fig. 9. Adult ratios per pentade at Foteviken (circles) and Löddesnäs (squares) in 1986. A "normal" year and both distributions conform, but the catch was unsuccessful at Löddesnäs in pentade 48 (only two catches, both half failures). The mean percentage of this pentade has been increased to 35.6 %, this value + true percentages has been applied to 1000 birds from pentades 46 - 51 (see Fig. 8). After recalculation the percentage will be equal = 16.8 % at both sites. During a second adult outbreak 9 - 11 September (3 catches with 72 2y+, 392 1y = 15.5 % 2y+) in Foteviken there was no catch at Löddesnäs.

age ratios 1992

Fig. 10. Adult ratios per pentade at Foteviken (circles) and Löddesnäs (squares) in 1992. There was no catch in Foteviken in pentade 49, no catch at Löddesnäs in pentade 47, these two percentages have been assessed (filled circle and square), finally these percentages + true percentages have been applied to 1000 birds from pentades 46 - 51 (see Fig. 8). Resultant values 20.4 % (Foteviken), 22.1 % Löddesnäs.

age ratios 1993 and 95

Fig. 11. Adult ratios per pentade at Foteviken (circles) and Löddesnäs (squares) in 1993 and 1995. In 1993 adult migration culminated early in the season, the regressions conform at both sites; both constant and regr. coeff. highly significant at Foteviken, n.s. at Löddesnäs. In 1995 catching numbers were poor at both sites, and there is no coverage in the central interval (pentades 48 - 49), the regressions are significant on the 5 % level. Percentages from the regressions have been applied to 1000 birds from pentades 46 - 51 at both sites (see Fig. 8); the values from 1993 (Foteviken 15.7 %, Löddesnäs 20.2 %) should be fairly correct, those from 1995 (11.4 and 18.5 % resp.) are on the low side but should give a correct indication of the general trend.

age ratios 1967-97

Fig. 12. Adult ratio (± 1 s.e.) of the total catch at Foteviken 1967 - 1995 (filled circles, dashed line) and Löddesnäs 1979-97 (open circles, solid line), corrected values shown without standard errors.

The survey of single years clearly shows, that there is a fair degree of freedom in the timing of migration even to adult birds; the peak of adult migration will range at least from the opening of pentade 48 (24 August) till pentade 51 (10 September), i.e. more than two weeks. Delays have three major causes: 1. events on breeding-grounds, 2. periods of warm weather, that will effectively quench the migration urge, and 3. periods of strong winds, connected with lows from the Atlantic. The last influence matters less to adults than to juveniles; adult "build-ups" in Foteviken often have coincided with strong winds in the NW sector, while 1y birds are more inclined to wait for moderate winds.

How are these fluctuations in age ratios reflected in population levels - in our case sampled by catching numbers? Fig. 13 shows the average catch per net (Table I) at Foteviken 1975 - 99 and at Löddesnäs 1979 - 1997. Neither site has a significant regression (coefficients of determination 0.02 (FOV) and 0.13 (LNÄ)) over the period of study, and the scale of the figure must not deceive; what it shows is an exceptional degree of stability in catching numbers; some 10 individuals per net and catching occasion at Foteviken, some 15 at Löddesnäs, which is the better site (higher density of roosting birds). Or, if normal distributions of values are assumed: Foteviken: 11.2 ± 0.6 %, s.d. 3.0 %; Löddesnäs: 17.7 ± 0.8 %, s.d. 3.1 % (note: other values than those calculated from all years pooled!). The introduction of tape lure at Foteviken in 1986 is most certainly reflected by the jump of the curve that year.

catchpernetandoccasion

Fig. 13. Mean catch per net and occasion (values from Table I) at Foteviken 1975 - 99 and at Löddesnäs 1979 - 97. Regressions are shown as a hint of possible trend, but neither is significant.

The continued investigation of the material presented in Figs. 6 - 12 must pass, step by step, from the general to the particular, in order to identify all sources of error in our sampling. A natural intermediate step is to calculate the correlation between age ratio and population level, the latter measured by the mean catch per net and occasion. The basic data are shown in Fig. 14:

TO BE CONTINUED HERE

Time distribution of different subspecies in the catch

From the very beginning of the project adult males belonging to the "subspecies" flava and thunbergi were noted separately. It soon turned out, however, that there are not two distinct, well-separated subspecies flava and thunbergi migrating by way of S. Sweden in autumn, but rather a continuous, unbroken cline from "flava" type to "thunbergi" type, with every conceivable intermediate occurring in fair numbers. Even within the flava group we could discern distinct types, and the first to be isolated was the "southern flava", the breeding Yellow Wagtail of the countryside surrounding Foteviken and Löddesnäs. This population is two-brooded - probably the only flava population laying two broods in Sweden - and it soon became evident, that all undeveloped juveniles (dark lines on head, band across breast, unmoulted coverts) ringed in August and September were 2nd brood birds belonging to this form. It has a very typical pastel hue to the whole adult plumage, and both male and female are characterized by a thick, creamy supercilium tending to broaden at both ends.

Unfortunately we did not note "southern" flava separately from the beginning, although we could recognize them; all "flava type" birds went under one single heading between 1967 and 1986. Instead the first step was to isolate the obvious hybrid flava/thunbergi; it was noted separately at Foteviken from 1967, and our own classification could take support and guidance from Sammalisto (1968). After that we gradually began to separate "southern" and "northern" flava, and these distinctions were put into practice in 1987 and 1988. In the latter cases we concentrated on adult males, females do not "suggest" their origin in the way males do. Finally, from 1988 onwards, we were confronted with green-headed and green-backed juveniles, strongly reminding of the subspecies flavissima, many of them even with yellow tone in their supercilium (1y birds of this subspecies are said to have buffish-white or white supercilium). Any remaining doubt was cleared away by the first adult male; up till and including 1994 there followed fifteen more adults, in addition 100 "possible" juvenile flavissima were ringed. In addition two adult males of feldegg type were caught, one at Foteviken 10.9.91, one at Löddesnäs 27.8.93. So, in our heyday, we discerned seven different main forms in South Swedish roosts, the type appearance of males in winter plumage is shown in Fig. 15:

Fig. 15. Major male types of Motacilla flava, discerned in S. Swedish roosts in autumn. Upper left: flavissima, centre: southern flava, right: standard flava. Middle left: northern flava, centre: hybrid (northern) flava/thunbergi, right: thunbergi, third row: aberrant thunbergi or feldegg. (Or maybe is mid row left more of true "hybrid" and mid row centre a good thunbergi; the latter category was included among thunbergi at Löddesnäs.) The female southern flava will in most cases have a very broad supercilium, too, while many thunbergi females lack supercilium and have very dark ear-coverts. The northern forms tend to be more saturated in colour than southern forms are.

Comment on the "feldegg" bird, depicting the individual from Foteviken 10.9.91: Sammalisto (1961) had 3 thunbergi males out of 2600 with entirely black heads (i.e. feldegg type), but feldegg is said to have no white in the lore region. Cf. Thorsten Stegmanns picture of a male in summer plumage in "naturfotogalerie" ! If this holds true of the autumn plumage as well - and this is by no means established - the bird shown belonged to the subspecies thunbergi - a memento for ringers and bird-watchers!

Pictures made in ColorIt after flashlight photos from Foteviken in September 1991.

Fig. 16 shows the distribution of the 100 possible juvenile and the 16 adult flavissima from Foteviken and Löddesnäs pooled - Figs. 17 - 23 the distributions of adults of the five other categories from Foteviken and of flava and thunbergi from Löddesnäs. Note that northern flava before 1986 and southern flava before 1988 are included in Fig. 18.

Fig. 16. Distribution of 16 adult and 100 possible juvenile flavissima at Foteviken and Löddesnäs 1988 - 94. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 17. Distribution of 182 adult southern flava from Foteviken 1988 - 99. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 18. Distribution of 2402 indiscriminate adult flava from Foteviken 1967 - 99. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 19. Distribution of 251 adult northern flava from Foteviken 1986 - 99. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 20. Distribution of 332 adult hybrids flava/thunbergi from Foteviken 1967 - 99. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 21. Distribution of 424 adult thunbergi from Foteviken 1967 - 99. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 22. Distribution of 1102 indiscriminate adult flava from Löddesnäs 1979 - 97. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

Fig. 23. Distribution of 308 adult thunbergi from Löddesnäs 1979 - 97. Median date shown with arrow; pentade 49 = 29.8 - 2.9.

TO BE CONTINUED

Till innehållsförteckningen för ärlor/To wagtail contents

Till innehållsförteckningen över morfologi/To morphology article contents

Tillbaka till startsidan/Back to start page