Moult of some Palaearctic Warblers Wintering in Uganda

Transcription

1 Bird Study ISSN: (Print) (Online) Journal homepage: Moult of some Palaearctic Warblers Wintering in Uganda D.J. Pearson To cite this article: D.J. Pearson (1973) Moult of some Palaearctic Warblers Wintering in Uganda, Bird Study, 20:1, 24-36, DOI: / To link to this article: Published online: 23 Jun Submit your article to this journal Article views: 140 Citing articles: 38 View citing articles Full Terms & Conditions of access and use can be found at

2 Moult of some Palaearctic Warblers Wintering in Uganda by D. J. Pearson Examination of the plumage of Palaearctic warblers as migrants or wintering birds at Kampala, Uganda, provided useful material for a consideration of moult in relation to the timing of spring and autumn migration. Reasons are suggested for observed interspecific differences in the moult pattern, and for the variation reported from different wintering areas in the African continent. THE PALAEARCTIC WARBLERS exhibit considerable variation in the timing of the main annual wing moult. Marked differences exist within individual genera, and in some cases species (Williamson 1967, 1968a, 1968b, Stresemann & Stresemann 1966, 1968), which appear to be closely linked with migratory behaviour. In general, birds which penetrate deep into tropical Africa delay replacement of their flight feathers until after autumn migration. From published information based on small numbers of museum specimens it is apparent that some warblers undergo an African moult in late autumn, whilst others commence late in winter and finish immediately before spring migration. In some cases there are considerable intraspecific differences in timing from one part of the African continent to another, and these will only be fully understood when more precise and detailed information is available from specific wintering localities. The present paper reports on plumage data for four species trapped during a two year study at Kampala, 0 20' N ' E. in southern Uganda. LOCALITY AND METHODS The Kampala area, on the north shore of Lake Victoria, provides a variety of habitats ranging from tropical forest, papyrus swamp and dense secondary bush to cultivation and scattered patches of grassland. Details of the status and migration of common Palaearctic visitors are described elsewhere (Pearson 1972). Four migrant warblers were trapped regularly over at least part of the northern winters 1966/67 and 1967/68. Reed Warblers Acrocephalus scirpaceus commonly inhabited dense moist thicket from late November to April. Sedge Warblers A. schoenobaenus were confined to dense waterside cover on the shores of the lake; they were scarce until midwinter, but became locally abundant between February and April. Garden Warblers Sylvia born and Willow Warblers Phylloscopus trochilus were common in scattered bush and forest edge habitats from October to April. All four were winter residents, but Garden Warblers and Willow Warblers also occurred abundantly on southward passage, returning in smaller numbers in spring. 24

3 MOULT OF SOME PALAEARCTIC WARBLERS Several sites in bush and tall grass cover were worked regularly with up to 100 metres of mist-net, at and just after dawn or during the late afternoon. The warblers caught were examined for moult and plumage wear. A considerable number were replacing wing and tail feathers; others were undergoing partial moult involving the body plumage. Warblers which renewed their wing and tail feathers at Kampala followed the typical sequences of passerine moult (Snow 1967). Replacement of primaries spanned the entire period of wing and tail moult except for the occasional early loss of a few greater coverts, and, in Willow and some Garden Warblers, the final stages of growth of the two inner secondaries. Following Newton (1967), the overall stage of wing moult has been recorded by reference to the primaries alone. For each bird, each new feather has been allotted a score of 1-5, according to its stage of growth, so that individuals with primaries in moult can be given a total score, P. of between 1 and 49. For Reed, Garden and Willow Warblers it was possible to determine the usual time of onset and completion of replacement of the primaries. Moult scores for these species have been plotted against date, and the slope used to deduce the mean duration of moult in the individual. For Reed and Garden Warblers mean dates have been calculated for successive groups of scores (1-10, etc.), but records for Willow Warblers being fewer, a slope has been fitted by eye. Although the great majority of the Acrocephalus warblers caught during winter showed no wing or tail moult, conclusions could be drawn from the state of wear of their plumage. PLUMAGE AND MOULT Reed Warbler Acrocephalus scirpaceus This species occurred in a variety of plumage states throughout the winter (Figure 1). Worn unmoulted birds were present from November (a few as early as October) until early February, but individuals in completely fresh plumage were trapped commonly from December (one as early as 18 November). Active moult of the primaries was recorded in every month from October to April, but was common only between late November and early March. By mid-april, the main time of spring departure, a number of birds had plumage already four or five months old, and were becoming slightly worn on wings and tail. Many of the unmoulted birds present in late autumn remained to renew their plumage at Kampala during the winter, but few of these were likely to have finished before February. Newly moulted birds, however, comprised 15 per cent of captures in early December, 41 per cent in late December and as many as 63 per cent in January. It was therefore concluded that many birds, probably the majority, renewed their plumage before reaching the area. The main arrival of unmoulted birds was during November and early December, and appears to have been followed by an influx of moulted birds a few weeks later, 25

4 BIRD STUDY Figure 1. Progress of moult in the Reed Warbler at Kampala, Uganda. Primary scores of moulting birds are plotted against date, and numbers of unmoulted birds (P = 0) and completely moulted birds (P=50) are shown grouped into five-day periods. Solid and dashed lines join records of the same bird in the same season. Closed circles: age indeterminate. Triangles: definitely adult. Large crossed open circles: mean dates for groups of ten moult scores. Practically all the unmoulted birds trapped in late autumn had worn primary tips. Some were adults, but the majority were considered to have been in their first year because: (a) the outer edges of their flight feathers and upper wing coverts were not unduly worn or faded; (b) their wing-formula tended to differ from that of freshly moulted birds in a manner not ascribable to abrasion, but in keeping with differences between adult and juvenile birds on British breeding grounds (personal observation); (c) whereas 112 birds in fresh plumage trapped in 1967/68 included 16 retraps from 1966/67, 67 unmoulted or moulting birds included none. 26

5 MOULT OF SOME PALAEARCTIC WARBLERS so so Sh. 20 LA; DEC JAN FEB MAR APR -13 Figure 2. The state of moult of Sedge Warblers trapped at Kampala, Uganda. Primary scores are plotted against date for moulting birds. Numbers handled in each half-month with P=0 and P=50 are shown. In the histograms for birds with P=50, the hatched portions represent birds in partial spring moult (body only) and the shaded portions those in partial spring moult involving tertials and/or tail feathers as well as the body. Birds which moulted before reaching Kampala were presumed to have included first year individuals as well as adults. There was wide variation in the date of onset of local moult, but calculated mean dates for successive groups of moult scores (Figure 1) indicate that the mean duration of primary moult in the individual was days. Several birds were caught more than once at different stages of moult and there appeared to be considerable variation in speed. Whilst one bird probably took less than 65 days, another apparently took over 80, and an intermediate duration was indicated in four other 'useful' retraps. Mean dates of onset and completion of moult were late December and early March respectively. 27 S.

6 BIRD STUDY Birds in wing moult were usually growing only one or two primaries per wing concurrently, and arrested moult was occasionally observed. Replacement of the secondaries commenced at a primary score of between 15 and 25, and was complete before full growth of the outer primaries. Tail moult commenced at about P=14 and usually finished between P=25 and 30; the six pairs of feathers were grown practically together. Replacement of the tertials, most of the wing coverts, and the body and head plumage occurred rather early in moult. Sedge Warbler Acrocephalus schoenobaenus Seven were caught during December. Of these, one (first-winter, but very worn) had not moulted, one (first-winter) had just commenced, and five were in fresh plumage. The main arrival of the species, however, was not until late winter and consisted almost entirely of moulted birds (Figure 2). In fact, many trapped in February already showed slight abrasion of wing and tail feathers judged to have been renewed during the previous autumn. A few birds which were still in mid-moult in late February (Figure 2) would have been unlikely to have finished much before the end of March. These were replacing one to three primaries per wing concurrently, and the relative timing of moult in different feather tracts appeared to be similar to that of Reed Warblers. Many birds trapped between February and April were moulting the feathers of the head, breast and mantle; a few were also replacing some of the tail feathers and tertials, but not the wing feathers. Birds undergoing this early spring partial moult tended to be those with most noticeably worn primaries and tail, i.e. those whose complete moult had presumably occurred earliest. Garden Warbler Sylvia born Wintering birds moulted completely. About two-thirds of late autumn captures had unabraded primaries, usually with pale tips, and were assumed to have been in their first winter. Until well into December they were easily separable from worn and undoubtedly adult birds. There was no indication that unworn autumn birds might have included adults which had moulted their flight feathers before migration (cf. Gladwin 1969); all six late autumn retraps from a previous season's ringing were considerably abraded. The great majority of captures between January and early March were replacing their primaries, but unmoulted birds were occasionally trapped to mid-january (one in early February) and a few completely moulted birds were recorded in February (Figure 3). The calculated mean dates for successive groups of primary scores indicate a more or less linear increase with time. They suggest a mean duration of individual primary moult of about 70 days. This figure is generally supported by the evidence available from retrapped birds; an average increase in score of at least 0.65 per day had occurred in three individuals over the first third of their moult. Assuming a duration of 70 days, and extrapolating backward and forward for each moulting bird, the mean dates of onset and completion were calculated as 28 28

7 Figure 3. Progress of moult in the Garden Warbler at Kampala, Uganda. Primary scores of moulting birds are plotted against date, and numbers of unmoulted birds (P=0) and completely moulted birds (P=50) are shown grouped into five-day periods. Solid and dashed lines join records of the same bird in the same season Closed circles: age indeterminate. Triangles: known adults. Squares: known firstwinter birds. Large crossed open circles: mean dates for groups of ten moult scores. The shaded portions of the histograms for birds with P =50 represent those with head moult incomplete. December and 8 March respectively. The fact that near these dates about 50 per cent of birds caught were in active moult implies that there was little trapping bias in favour of non-moulting individuals. The age of a number of moulting birds was known from their retrap history and it was concluded that adults tended to commence slightly before first-winter birds. The mean extrapolated starting date for seven moulting retraps 29

8 BIRD STUDY r JAN FEB MAR Figure 4. Progress of moult in the Willow Warbler at Kampala, Uganda. Primary scores of moulting birds are plotted against date, and numbers of unmoulted birds (P=0) and completely moulted birds (P=50) are shown grouped into five-day periods. The thick solid line representing mean moult progress has been drawn by eye. Other solid lines and dashed lines join records of the same birds in the same season. Closed circles: age indeterminate. Squares: first-winter birds. known to have been adults was thus 15 December, whereas that of five known to have been first-winter was 30 December, significantly later (P<0.05). Moreover, of birds retrapped from the previous autumn, all five still in old plumage after 20 December were young birds, whilst the only two with wing moult complete before mid-march were adults. 30

9 MOULT OF SOME PALAEARCTIC WARBLERS After the initial stages of replacement, the number of primaries in growth in each wing was usually two, sometimes three. Moult of the secondaries, which commenced at P=18-25, was complete just before or just after growth of the leading primaries. Tail replacement commenced at P=20-25 and was complete at P=35-40; as in the Reed Warbler, the new feathers developed practically together. The tertials and most of the wing coverts were replaced rather early in moult, but the body plumage mainly during the later stages. Head moult occurred last, and was still in progress in most birds trapped in March, including a number with wing moult complete. Willow Warbler Phylloscopus trochilus All wintering birds underwent complete moult. Primary moult was recorded from mid-december to early April, and was active in nearly all birds caught between January and early March. There was no evidence that its timing in young birds (usually to be distinguished by their much yellower under parts) differed from that of adults. The species was only occasionally caught in mid-winter, when least abundant, so that records of active moult are relatively few. However, the line in Figure 4, fitted by eye, suggests a mean date of onset during the first week of January. Two 'useful' retraps indicated that moult in the birds concerned was completed in about 74 days and within 77 days respectively. A third, caught three times during the late stages was perhaps moulting more slowly. Completely moulted birds trapped during the third week of March were assumed to have included early passage migrants; the species increased in abundance about this time. The number of primaries in active replacement was sometimes three per wing (early in the moult), but more often one or two. As far as could be ascertained, replacement of the secondaries usually commenced at.p=20-30 (rather later than in most Reed and Garden Warblers), growth of the two innermost feathers being completed after that of the outer primaries. The tail feathers were moulted centrifugally in rapid succession, starting usually at about P=25 and finishing at about P=35. Active replacement of body and head plumage was evident at all stages of wing moult. MOULT IN RELATION TO MIGRATION Garden and Willow Warblers underwent complete moult in Kampala winter quarters and departed in spring with fresh flight feathers. Most of the Acrocephali resident in late winter, on the other hand, were already in fresh plumage on arrival, and were beginning to look rather worn by the time of spring migration. They made a very late appearance and presumably moulted elsewhere in northeast Africa. Figure 5 shows the timing of the Kampala moult of the Reed, Garden and Willow Warblers relative to periods of migration, as ascertained by ringing and observation over three seasons. Mean date of onset and individual duration were similar in the three species, but variation in 31

10 1 MAIN 5" ARRIVAL 5,' OF 5 UNMDU LT E 0 5 BIRDS 5,' MAIN ART UM N ARRIVAL MAIN SPRING PAS SAGE 5 AND 5 DEPART UR MAIN AUTUMN WILLOW ; PASSAGE WARBLER.; AND ARRIVAL X I MAIN % SPRING 5 PASSAGE AND e DEPARTUR IV Figure 5. Moult and migration seasons of three migrant warblers at Kampala, Uganda. The onset (left) and completion (right) of wing moult are shown as percentages by half-monthly periods. timing was greater in the Reed Warbler (95 per cent spread, more than three months) than in the other two (95 per cent spread, about two months). Most locally moulting Reed Warblers appear to have commenced several weeks after arrival and finished within six weeks of departure. Most Garden Warblers commenced some weeks after reaching the area, when the majority of autumn birds bound farther south had 32 II

11 MOULT OF SOME PALAEARCTIC WARBLERS departed. They finished about four to six weeks before northward migration, with head moult continuing for a further week or two. Willow Warblers appeared to moult even later relative to migration. Most commenced some two months after the peak passage and arrival period, and finished growth of the inner secondaries within three weeks of spring departure. Long distance migrants gain two obvious advantages from delaying wing moult until arrival in the tropics. Firstly, their flight feathers are in a relatively fresh condition for the spring migration, which tends to be more rapid and demanding than the autumn journey. Secondly, there is more time available for moult in winter quarters than in the few weeks between completion of breeding and commencement of autumn migration. Most warblers wintering at Kampala were of eastern origin (Pearson 1972), but it is instructive to compare duration of wing moult with that in similar species replacing their flight feathers near British breeding grounds (Table I), estimated from records accumulated by the British Trust for Ornithology Moult Enquiry. Replacement of primaries tends to proceed at a much more leisurely tempo in winter quarters than near the breeding grounds, and this would seem to be due to the fact that fewer feathers are grown together, rather than to a slower rate of growth of individual feathers (Table I). In the Willow Warbler, apparently unique amongst Palaearctic warblers in undergoing two complete moults per year (Williamson 1967), the estimate of days for the mean duration of primary replacement in Uganda (presumably typical of the African moult) makes an interesting comparison with the much shorter period of days required in Britain, and probably typical of the post-nuptial moult in Eurasia. Garden Warblers which breed in southwest Europe sometimes moult their flight feathers before autumn migration (Gladwin 1969), but the number of primaries undergoing concurrent replacement is apparently greater than during moult in winter quarters, and retrap data given for two British moulting birds indicate a much shorter duration than the 70 days usual in Uganda. TABLE I. COMPARISON OF THE DURATION OF WARBLER MOULT IN UGANDA WITH THAT NEAR ENGLISH BREEDING GROUNDS Species Usual duration of Mean no. of primaries primary moult in growth at once (days) per wing* Wing moult'in Reed Warbler (53) 1.8 (23) Kampala winter Garden Warbler (102) 2.4 (51) quarters Willow Warbler (22) 2.0 (12) Wing moult Blackcap (100) 3.5 (34) near breeding Whitethroat (200) 3.6 (104) grounds in Lesser Whitethroat (25) 4.5 (2 only) southern Willow Warbler (200) 3.8 (87) England Chiffchaff (60) 3.4 (23) *Between complete growth of the first primary and loss of the ninth. NOTE: The numbers of moulting birds on which the figures in each column are based are given in brackets. Figures for moult duration in southern England were estimated from the slope produced by plotting mean dates for various groups of moult scores. 33

12 BIRD STUDY COMPARISON WITH OTHER WINTERING POPULATIONS The timing of Garden Warbler and Willow Warbler moult at Kampala was in good agreement with that reported for East African birds by Williamson (1967, 1968a). Willow Warblers which winter in eastern Africa (mostly P. t. acredula) tend to moult later than those of western Africa (P. t. trochilus) but this presumably corresponds with the later spring migration and breeding season of the northern race. Garden Warblers tend to moult a few weeks earlier in eastern than in other areas of the continent, and this is less easy to explain. It could perhaps indicate the presence of a high proportion of relatively early breeders from southwest Asia. Reed and Sedge Warblers exhibit marked differences in timing from one wintering area to another, and tend to present a bewildering variety of moult states within individual populations. An attempt is made below to relate moult findings at Kampala to information from other wintering areas, based mainly on the statements of Williamson (1968b), and on examination of the wintering and passage specimens in the collections of the British Museum (Natural History). The Reed Warbler rarely winters south of latitude 15 0 S. (Vaurie 1959), and there is little available museum material or published information with which to compare Kampala observations. In the northern tropics, moult has usually been recorded in late autumn. The few examples mentioned by Williamson suggest that it commences later where it occurs farther south, and indeed most birds moulting at Kampala finished late in February or early in March. However, the great majority of specimens from late December onwards from all parts of the winter range are in completely fresh plumage. In southwest Uganda, as at Kampala, many birds arrive moulted (M. P. L. Fogden, pers comm.), and in southern Kenya the great majority of birds appear in fresh plumage in late December and January (personal observation). Most birds which migrate to the equator and beyond evidently pause to renew their plumage on the way south. Sedge Warblers wintering in southern Africa usually moult in late winter and early spring, commencing some time after their arrival. Thus, November and December specimens from South Africa, Rhodesia and Botswana are unmoulted, whilst those collected between January and March are mostly in wing moult, with April birds finishing or in very fresh plumage. The early moult timing of the great majority of Kampala birds thus contrasts with the southern African picture, but is probably typical of equatorial wintering populations. Specimens from other parts of east and west tropical Africa, south to about latitude 50 S. show evidence of a late autumn moult, practically all from early winter being in fresh plumage and most from early spring having somewhat worn primary tips. Available material suggests that populations wintering between about 5 S. and 15 S. commonly include both early and late moulters. 34

13 MOULT OF SOME PALAEARCTIC WARBLERS Several late autumn specimens from the Sudan are actively renewing their flight feathers. J. S. Ash (in litt.) has observed moult in Ethiopia during November, but there appears to be little evidence that it occurs at this time farther south. Like those of Kampala, most southern Kenyan wintering birds make their appearance in fresh plumage during the winter months (personal observation). Indeed, it would seem, as in the Reed Warbler, that most early moulting birds with final winter quarters in equatorial or southern tropical latitudes must renew their plumage before arrival there. DISCUSSION Length of migration and final wintering latitude profoundly influence the timing of moult in trans-saharan passerine migrants. Delay of the main moult until after arrival in the tropics tends to occur in species or populations which reach equatorial or southern latitudes. For example, the Garden Warbler and the eastern race of the Whitethroat Sylvia coinmunis icterops (Stresemann & Stresemann 1968) perform longer migrations than other Sylvia warblers, and winter mainly south of the equator. They are the only members of the genus which moult their flight feathers in winter quarters. Interestingly, the southwestern breeding Garden Warblers in which moult before autumn migration has been observed probably migrate much less far than other populations. Within genera or species which moult completely in Africa, there is a tendency for birds which migrate farther south to moult latest in winter, so that their flight feathers are fresh immediately before spring migration. This is evident within the genera Locustella, Hippolais and Lanius, and in Acrocephalus, where there are intraspecific as well as interspecific timing differences. The Marsh Warbler A. palustris, with final wintering grounds in southeast Africa, moults during February and March (Williamson 1968b). The Reed Warbler, on the other hand, winters near the equator and moults mainly in late autumn; some birds reach, and migrate beyond, the equator in old plumage and moult during the winter months, but not as late as most Marsh Warblers. The majority of Sedge Warblers and Great Reed Warblers Acrocephalus arundinaceus arundinaceus 1 zarudnyi (Pearson, in press) moult early during their stay in Africa, but the southernmost populations of these two species renew their plumage shortly before spring migration. Many areas immediately south of the Sahara must provide an abundance of insect food after the summer rains This is presumably exploited during the autumn months by passerine migrants, many of which migrate to spend the latter part of the winter much farther south. A number of species reach the equator many weeks after their passage through the Middle East. For example, few Sylvia and Acrocephalus warblers appear in southern Uganda before November (Pearson 1972), and most Sprossers Luscinia luscinia, Marsh Warblers and Whitethroats pass through eastern Kenya during late November and early December (G. C. Backhurst and D. J. Pearson, unpublished observations). Some species appear to use this 35

14 BIRD STUDY pause in the northern tropics to renew their plumage. Thus, many of the Reed, Great Reed and Sedge Warblers with final wintering grounds in eastern and southern tropical Africa arrive already moulted between December and February, and many Olivaceous Warblers Hippolais pallida appear in eastern Uganda in fresh plumage at the end of the year (D. J. Pearson and J. G. Rolfe, unpublished observations). It would seem that many passerine migrants have at least two widely separated areas in Africa where they are resident for long periods of the non-breeding season. The term 'wintering area' applied to the southernmost of these alone is somewhat misleading. It is not necessarily the African moulting ground, and may not be reached until February. ACKNOWLEDGEMENTS I am extremely grateful for the use sf the facilities of the British Museum (Natural History). I also wish to express my thanks to Dr J. H. Phillips and Dr I. M. Smith for valuable field assistance in Uganda, and to those who have contributed British warbler records to the B.T.O. Moult Enquiry. SUMMARY An account is given of the plumage state and moult of some 1,100 Palaearctic warblers trapped at Kampala in southern Uganda. Wintering Garden and Willow Warblers moulted completely, most birds commencing in late December or early January and finishing in March. Most Sedge Warblers, on the other hand, arrived in fresh plumage, having completed moult elsewhere in late autumn. Wintering Reed Warblers comprised two groups. One, consisting mainly of young birds, arrived in worn plumage and underwent moult in the area; the other arrived in fresh plumage during December and January. Garden Warblers, Willow Warblers and most locally moulting Reed Warblers finished shortly before spring migration, and departed with their plumage in fresh condition. The majority of the wintering Acrocephali, however, commenced spring migration with their flight feathers already slightly worn. Sedge Warblers (but not Reed Warblers) commonly replaced much of the body plumage in early spring. The usual duration of primary moult was estimated at days in the Garden Warbler, days in the Reed Warbler and days in the Willow Warbler. Thus moult in Ugandan winter quarters is much more leisurely than post-nuptial moult on breeding-grounds which occupies days in the Willow Warbler. REFERENCES GLADWIN, T. W Post-nuptial moult in the Garden Warbler. Bird Study, 16: NEWTON, I Feather growth and moult in some captive finches. Bird Study, 14: PEARSON, D. J The wintering and migration of Palaearctic passerines at Kampala, southern Uganda. Ibis, 114 : PEARSON, D. J. (In press). East African observations concerning the timing of moult in the Great Reed Warbler. SNOW, D. W A Guide to Moult in British Birds. British Trust for Ornithology Field Guide No. 11. STRESEMANN, E. and v Die Mauser der Vogel..1. Orn., 107, suppl. STRESEMANN, E. and V Winterquartier und Mauser der Domgrasmiicke Sylvia communis. I. Orn., 109: VAURIE, C Birds of the Palearctic Fauna. London. WILLIAMSON, K. 1967, 1968a, 1968b. Identification for Ringers (Revised Editions). 2. The Genus Phylloscoups. 3. The Genus Sylvia. 1. The Genera Cettia, Locustella, Acrocephalus and Hippolais. British Trust for Ornithology. Oxford. D. J. Pearson, Department of Medical Biochemistry, University of Nairobi, Box 30197, Nairobi, Kenya. 36