Site Map ; Contributors

Habitat use, moult and biometrics in the Manchurian
Reed Warbler Acrocephalus tangorum wintering in
Thailand

By Philip D. Round & Stephen J. Rumsey

ABSTRACT
Manchurian Reed Warblers Acrocephalus tangorum wintering at Khao Sam Roi Yot, Thailand, were largely restricted to mature Phragmites reeds. The mean wing length (± sd) of 105 Manchurian Reed Warblers was 54.5 ± 1.41 mm. There was no evidence of differences in body weight in spring, autumn or winter. Both adults and first-winter birds underwent a complete moult soon after arrival in their winter quarters. Primary moult duration was estimated to be 59 days. Moult of secondaries was suspended in approximately one-third to one-fifth of birds, the few retained, old, unmoulted, feathers usually being replaced later in the winter. No further moult was usually detected prior to northward spring migration.

INTRODUCTION
The Manchurian Reed Warbler Acrocephalus tangorum has a localised breeding distribution in the northeast Chinese province of Heilongjiang and the Nei Mongol Autonomous Region (Cheng 1987, Alström et al 1991), and in the Russian Far East (Shibnev & Gluschenko 1977, Stepanyan 1978, Gluschenko 1981, 1989). Its wintering grounds were unknown until a population was discovered in 1981 in a reedswamp at Khao Sam Roi Yot, Prachuap Khiri Khan Province, Thailand (Round 1993). The species is now known to be more widespread in southeast Asia. In addition to a few sightings elsewhere in Thailand (Round 1993, Round & Jukmongkol 2001, 2002), it is now known to winter in Cambodia (BirdLife International 2001, Davidson 2001, Robson 2000), southern Laos (Round 1998), and Hong Kong (Leader & Lewthwaite 1996, Carey et al 2001). On passage, it has been recorded from northeast China (BirdLife International 2001, Brazil 1992, La Touche 1912, Williams 2000), Hong Kong (Carey et al 2001) and northern Vietnam (Tordoff & Eames 2001).

Though the Manchurian Reed Warbler was formerly considered to be conspecific with Paddyfield Warbler A. agricola (Vaurie 1959) or even Black-browed Reed Warbler A. bistrigiceps (Williamson 1976), mitochondrial DNA evidence indicates that it is better treated as a distinct species more closely allied to Blunt-winged Warbler A. concinens (Leisler et al 1997). Because of its small breeding and wintering range, and therefore presumably small world population, it is considered to be globally threatened, and is listed as ‘vulnerable’ inthe 2000 International Union for Conservation of Nature and Natural Resources (IUCN) Red List (BirdLife International 2001).

This paper describes the habitat choice and moult cycle of the Manchurian Reed Warbler wintering at Khao Sam Roi Yot, Thailand, and makes general observations on other Acrocephalus warblers encountered there.

METHODS

Study area
Khao Sam Roi Yot National Park is situated in Prachuap Khiri Khan Province, Thailand, between 12°05’ to 12°20’N and 99°52’ to 100°02’E. It encompasses part of a reedswamp and lowland marsh, bounded on its eastern, coastal, margins by steep limestone outcrops. To the west, it grades into paddies, scattered stands of Borassus palms, and other cultivated and settled land. The least disturbed part of the 50 km2 marsh is dominated by a large (4-6 km2) stand of reeds, Phragmites karka. Extensive areas of reedmace, Typha angustifolia, are also present, especially around the disturbed margins. The marsh has been subject to a low to moderate level of human use for many decades, principally for fishing and catching waterfowl, and the remains of old ditches and field systems are evident in some parts. Since 1986, it has suffered severe degradation, with encroachment to establish prawn farms, and plantations of Eucalyptus and Casuarina.

Mist-netting, ageing and moult
Acrocephalus warblers were caught using mist-nets at Khao Sam Roi Yot during nine netting and ringing sessions between April 1995 and October 2002. On the first session, which lasted 19 days, nets were set in a variety of different vegetation types around the edges of the reed-swamp, to investigate habitat preferences by Acrocephalus warblers. The vegetation types were mature Phragmites, among which tall, woody stems were frequent; young Phragmites, found usually in marginal situations, where mixed with Typha angustifolia; pure stands of T. angustifolia, a stand of Scirpus sp, and scrub. Nets were usually set over open water, or along the edges of ditches, where embankments permitted access. Based upon the findings of this initial survey, all subsequent netting sessions, totaling 36 days, concentrated preferentially in mature Phragmites. Most netting was carried out between 06.00 h and 11.00 h. All trapped birds were identified, ringed, measured and examined for moult and feather wear. Birds were weighed to the nearest 0.1 g using Pesola spring balances.

First-year Manchurian Reed Warblers in early autumn were distinguishable from adults by their relatively fresh remiges and rectrices, although by November the two age groups are thought to be indistinguishable. Juveniles and first-years of some other Acrocephalus species show a dull, dark brown or slightly grey-tinged iris in comparison with adults, in which the iris is more reddish or chestnutbrown (Karlsson et al 1988). Iris colour was tentatively used to age birds from November 2000 onwards, though was not consistently used for ageing in years prior to this. This is not, however, a completely reliable character for all Acrocephalus species. One Manchurian Reed Warbler in the very early stages of a full moult was clearly an adult due to its worn body plumage and primaries, although it had irides which lacked warm brown tones. Similarly, an Oriental Reed Warbler A. orientalis, retrapped in at least its third calendar year, still showed a grey-tinged iris. It is possible that birds with warm brown irides are more than one year old, but that not all birds with dull brown or grey-tinged irides are necessarily in their first-year.

Moult was recorded using the method described by Ginn & Melville (1983); with the primaries numbered descendantly, each flight feather is given a score from 0 (old) to 5 (fully grown and new). In this account, the term complete moult is used to refer to moult involving both flight feathers and contour feathers, and does not necessarily imply that moult was completed without being suspended or arrested.

RESULTS
A total of 106 Manchurian Reed Warblers was caught during the nine trapping sessions: 3–21 April 1995 (63 birds); 29 October–24 November 1996 (12 birds); 4–5 November 2000 (one bird); 30–31 December 2000 (six birds); 17–18 March 2001 (16 birds); 30 April–1 May 2001 (one bird); 22–23 October 2001 (three birds); 21– 22 September (three birds) and 6 October 2002 (one bird). The smaller catches in the autumn were partly due to extensive seasonal flooding which limited access to the reedbed. The April 1995 catch also included 128, 111, 18 and five Thick-billed Warblers A aedon. For all other periods combined, other species trapped included 110 Oriental Reed Warblers, 85 Blunt-winged Warblers and nine Black-browed Reed Warblers.

Habitat use
The expected numbers of Acrocephalus warblers were calculated for each habitat on the basis of the number of metre-hours (mh) of mist-netting carried out, assuming no preference among different habitats. The results for all habitats other than Phragmites were pooled. All four Acrocephalus species showed a strong, positive association with Phragmites reeds, and an avoidance of Typha angustifolia or mixed Typha and young Phragmites (Table 1; Black-browed Reed Warbler
2 = 152.83, P < 0.01; Oriental Reed Warbler
2 = 41.49, P < 0.01). Sample sizes of Manchurian Reed Warbler and Blunt-winged Warbler were too small to test statistically, but the distribution of both species was heavily skewed towards stands of mature Phragmites (Table 1). Elsewhere, where Manchurian Reed Warblers were netted in Phragmites, open water was usually present in ditches. A common behaviour of Manchurian Reed Warbler, when feeding, was to climb to the tops of tall reed stems with the tail cocked. This behaviour was never seen in Black-browed Reed Warbler, which tended to feed lower in the vegetation column.

No trapping was carried out in the peripheral areas of the marsh dominated by Eleocharis dulcis, where vegetation was too low for mist-netting, however, observations revealed little use of Eleocharis by any species of Acrocephalus warbler. A few Black-browed Reed Warblers were occasionally observed in Eleocharis close to the ecotone with taller vegetation.

Table 1 : Distribution by habitat of Acrocephalus warblers trapped in marshy areas of Khao Sam Roi Yot, Thailand, during April 1995. Total trapping effort was 14,498 m of net hours (mh).

Moult
Some adult, and possibly also first-year, Manchurian Reed Warblers undergo a partial moult, involving body feathers, and sometimes tertials and tail, before migration. Three adults caught on 21–22 September had mainly new, or a mixture of old and new, body feathers, but had not yet commenced moult of wing feathers. One had asymmetrically replaced eight rectrices and the innermost tertials on both wings, another had renewed all tertials, and the third had started moulting greater coverts, but no moult was observed in other feather tracts. Three individuals caught on 22 October, two of which were aged as first-years, appeared to have renewed the tertials before migration, since these feathers appeared significantly fresher than most other contour feathers, yet were not as new as the feathers then being moulted in.

The earliest instance of primary moult recorded was on 6 October, a retrap, first handled on 22 September, which had started to renew the two innermost primaries (moult score 3). Another adult, trapped on 6 October, had not yet started moult.

All other Manchurian Reed Warblers caught in October and November, including birds aged both as adults and first-years, were either in active moult of flight feathers, or had recently completed moult. The earliest birds to have completed primary moult were caught on 21 November, with two more on 24 November. Linear regression of primary moult score against date gave an estimate of 59 days for the duration of primary moult for the population (Fig 1). Birds completing moult by 21 November, the earliest date found during this study, should therefore have commenced moult around 23 September. Regressions of moult score against date can be used to estimate of duration of moult for the population, rather than for individuals (Ginn 1975, Pimm 1976), so that even for the early-completing birds, primary moult may not necessarily have started as early as this. The latest moulting birds might not complete moult until early to mid-December.

Moult patterns of rectrices appeared to be variable. The earliest had already dropped three central pairs of rectrices at primary moult score 3 and the first completely new tail was observed at moult score 40 (Fig 2). Three birds with primary moult scores of 20–26 were in early stages of growing all six pairs of rectrices, while another observed in the field, on 18 November 2000 also lacked a tail, indicating that some individuals may moult all tail feathers simultaneously.

The middle tertial was dropped at primary moult score 5 (one bird) and all new, fully grown tertials were already present at score 20 in two birds (still growing in two others, moult score 24 and 26). Secondary moult began at primary moult score 15–20 (two birds) and appeared to be completed coincident with completion of primary replacement. Three birds caught on 22 October with primary moult scores of 15, 20 and 24 were about midway through body moult, but a bird caught on 4 November with a score of 26 was already close to completing body moult. The sample was too small to give a more accurate description of moult sequence among feather tracts.

Moult of secondaries was sometimes suspended and resumed later in the winter, as summarised in Table 2. Seventeen of 80 spring-caught birds (21%) showed a mixture of old and newer flight feathers indicating some replacement during mid- or late winter.

It appears that in the Manchurian Reed Warbler, the general pattern is for no further moult to take place before spring migration. A few birds netted in March and April had some feathers in pin on the head, upperparts and underparts, and two individuals were replacing some tertials (Table 2). None showed extensive renewal of body feathers and all appeared strikingly worn. In addition, four specimens in The Natural History Museum, Tring, collected from Qinghuangdao, China, during late May to early June (Round 1993) were also extremely worn and faded; their condition appeared consistent with birds which had not recently undergone any pre-breeding body moult. Ironically though, the first record of Manchurian Reed Warbler from Khao Sam Roi Yot, on 6 May 1981 (Field number ACW 14, Thailand Institute of Scientific and Technological Research, Bangkok), involved an apparently unusual bird which had undergone an extensive moult of body feathers, tertials, tail feathers, and secondary six on one wing (Round 1993).

Table 2. Pattern of replacement of secondaries and tertials among Manchurian Reed Warblers handled in midwinter and spring at Khao Sam Roi Yot, Thailand. For a further 67 birds handled during this period, moult was complete with no arrest, or suspension, and subsequent flight-feather replacement.

X: feathers of the previous generation not replaced during the autumn/winter moult, 0: feathers replaced during complete moult in autumn/ early winter, N: feathers replaced during subsequent partial moult later in the winter, G: feathers in active growth.

Table 3. Body weights of Manchurian Reed Warblers netted at Khao Sam Roi Yot, Thailand. There were no significant differences in body weights between the three periods (Kruskal-Wallis test, P > 0.1).

DISCUSSION
We found a strong, positive association between Manchurian Reed Warblers and stands of mature Phragmites. This preference was found in all Acrocephalus species netted at Khao Sam Roi Yot. Although, in this study, the Manchurian Reed Warbler showed a clear preference for Phragmites, small numbers nevertheless utilise other swamp vegetation, especially in the absence of Phragmites. There have been a few sightings or captures in Typha around Bangkok (Round &
Jukmongkol 2001, 2002), low sedge in southern Laos (Round 1998), while in Cambodia, the species has been found in a wide variety of wetland habitats, including sedge beds, Sesbania scrub around pools in open, dry dipterocarp woodland, grassland and scrub mosaic and, especially, stands of tall grasses in the Tonle Sap inundation zone (Davidson 2001, P. Davidson pers comm). The extent of apparently suitable grassland habitat in the Tonle Sap inundation zone is vast (P.
Davidson, pers comm) and even if densities in grasses other than Phragmites should prove to be lower, the Manchurian Reed Warbler may be more widespread in winter than previously thought.

Linear regression methods applied to moult score, though somewhat imprecise (Seel 1976, Summers et al 1983, Underhill & Zucchini 1988), remain the most widely used method for estimating moult duration among passerines, so that the estimate of 59 days obtained for duration of primary moult in the Manchurian Reed Warbler has useful comparative value. Among species which moult on the wintering grounds, estimates of primary moult durations are 65–80 days for Eurasian Reed Warbler A. scirpaceus (Pearson 1973, in litt); 65 days for Sedge Warbler A. schoenobaenus (Ginn & Melville 1983); 35–91 days for individual Sedge Warblers in West Africa which were retrapped at least once during moult, and 42–51 days for a single Great Reed Warbler A. arundinaceus (Bensch et al 1991); and 50–60 days for Pallas’s Grasshopper Warbler Locustella certhiola (Nisbet 1967).

The Manchurian Reed Warbler shows some variability in the extent of the moult following arrival in the winter quarters. Although all individuals, from our sample, underwent a complete moult of primaries and rectrices, one-third to one-fifth failed to complete moult of secondaries. Most of those suspending moult replaced these older, more worn feathers later in the winter, and in spring showed two, or even three, ages of flight feathers. A few had a near complete moult in autumn
without undergoing any further major replacement of feathers (arrested moult). We found no consistent prebreeding moult of body and covert feathers in the birds before spring departure.

The Manchurian Reed Warbler shows some variability in the extent of the moult following arrival in the winter quarters. Although all individuals, from our sample, underwent a complete moult of primaries and rectrices, one-third to one-fifth failed to complete moult of secondaries. Most of those suspending moult replaced these older, more worn feathers later in the winter, and in spring showed two, or even three, ages of flight feathers. A few had a near complete moult in autumn
without undergoing any further major replacement of feathers (arrested moult). We found no consistent prebreeding moult of body and covert feathers in the birds before spring departure.

There was some evidence to suggest that tail feathers were dropped in rapid succession or simultaneously. This strategy is known in some other skulking, grassland or reedbed-inhabiting birds, such as Pallas’s Grasshopper Warbler (Nisbet 1967), some River Warblers Locustella fluviatilis wintering in Africa (D.J. Pearson, pers comm) and some Bradypterus warblers (pers obs). It probably occurs where impaired flight capability for short periods may not matter, presumably because the birds feed mainly by walking or hopping through reeds and dense scrub, without having to cross wide gaps.

The wing and tail feathers of reed-inhabiting birds often show more wear than those of species from other habitats, possibly due to frequent, direct abrasion from the dense habitat. This might place a strong selective pressure on feather replacement, provided that the food resources are sufficient to allow it. Manchurian Reed Warblers arrive at Khao Sam Roi Yot towards the end of the southwest monsoon, when vegetation is luxuriant, annual flooding in the marsh is at its peak, and insect
food is probably abundant. Conditions in the succeeding dry season, as water levels recede, probably vary from year to year, depending on the extent of the previous flooding. One interpretation of the variability of moult may be that the birds are adapted to take best advantage as near completely as possible when conditions are wettest and subsequently replacing as many previously unmoulted feathers as possible in years when food is abundant. Although Oriental Reed Warbler is
exceptional in that both adults and first-years moult completely on the breeding grounds, moult patterns of some other Asian-wintering Acrocephalus species seem to show some parallels with Manchurian Reed Warbler. Both Paddyfield Warbler and Blyth’s Reed Warbler A. dumetorum apparently either suspend moult, or moult some body feathers later in the winter after undergoing a complete moult in the early winter (Svensson 1992). Black-browed Reed Warbler, on the other hand, moults on the breeding grounds and undergoes no further replacement of flight feathers during winter (Medway
& Wells 1976, Williamson 1976). In spite of this, the flight feathers of Black-browed Reed Warblers in spring are scarcely more worn than those of Manchurian Reed Warblers, and may even appear less so. This may be related to differences in the foraging behaviour, combined with the shorter, broader tail of Black-browed Reed Warblers, which may be less susceptible to wear. One apparent paradox of the strategy of moulting early in the winter, following migration, as shown by
Manchurian Reed Warbler, is that both autumn and spring migrations are undertaken with worn flight feathers, when we might expect there to be a high premium for efficiently functioning feathers.

Additional trapping of birds throughout the nonbreeding period, particularly in early autumn and late spring, would further elucidate aspects of the annual cycle of the Manchurian Reed Warbler. In particular, we need to determine arrival and departure dates, improve estimates of the duration of moult for different age classes, and to understand the extent of occurrence and variability of any partial post-breeding and prebreeding moult of contour feathers.

REFERENCES
Alström, P., Olsson, U. & Round, P.D. (1991). The taxonomic status of Acrocephalus agricola tangorum. Forktail 6, 3–13.

Bensch, S., Hasselquist, D., Hedenström, A. & Ottoson, U. (1991). Rapid moult among palaearctic passerines in West Africa— an adaptation to the oncoming dry season? Ibis 133, 47–52.

BirdLife International (2001). Threatened birds of Asia: the BirdLife International Red Data Book. BirdLife International, Cambridge.

Brazil, M. (1992). The birds of Shuangtaizihekou National Nature Reserve, Liaoning province, China. Forktail 7, 91–124.

Carey, G., Chalmers, M.L., Diskin, D.A., Kennerley, P.R., Leader, P.J., Leven, M.R., Lewthwaite, R.W., Melville, D.S.,
Turnbull, M. & Young, L. (2001). The Avifauna of Hong Kong. Hong Kong Birdwatching Society, Hong Kong.

Cheng, Tso-Hsin (1987). A synopsis of the avifauna of China. Beijing, Science Press.

Davidson, P. (2001). A further twelve new species for Cambodia. Cambodia Bird News 7, 26–35.

Ginn, H.B. (1975) The timing and sequence of the complete annual moult in the Dunnock (Prunella modularis) in Britain over an eleven year period. Journal of Ornithology 116, 263–280.

Ginn, H.B. & Melville, D.S. (1983) Moult in Birds. BTO Guide 19. BTO, Tring.

Gluschenko, U.N. (1981) Nesting birds of the Khanka Lowlands. In Rare birds of the Far East (ed Litvinenko, N.M.) pp 25-33 Far East Science Centre, Academy of Sciences of the USSR (In Russian).

Gluschenko, U.N. (1989) Acrocephalus agricola tangorum. In Rare vertebrates of the Soviet Far East and their protection (eds P.A. Ler et al). pp 158–159. Nauka, Leningrad. (In Russian).

Karlsson, L., Persson, K. & Walinder, G. (1988) Ålderbestämning av rörsångare Acrocephalus scirpaceus med hjälp av irisfärg, tarsfärg och tungfläckar. Vår Fågelvärld 47, 141–146. (In Swedish, English summary).

La Touche, J.D.D. (1912) Acrocephalus tangorum. Bulletin of the British Ornithologists’ Club 31,10–11.

Leader, P.J & Lewthwaite, R.W. (1996) Manchurian Reed-Warbler: the first records for Hong Kong. Hong Kong Bird Report 1995, 119–122.

Leisler, B., Heidrich, P., Schulze-Hagen, K. & Wink, M. (1997) Taxonomy and phylogeny of Reed Warblers (Acrocephalus) based on mtDNA sequences and morphology. Journal of Ornithology 138, 469–496.

Medway, Lord and Wells, D.R. (1976) Birds of the Malay Peninsula. Vol 5. Witherby & Penerbit Universiti Malaya, Kuala Lumpur.

Nisbet, I.C.T. (1967) Migration and moult in Pallas’s Grasshopper Warbler. Bird Study 14, 96–103.

Nisbet, I.C.T. & Medway, Lord (1972) Dispersion, population ecology and migration of Eastern Great Reed Warblers Acrocephalus orientalis, wintering in Malaysia. Ibis 114, 451–494.

Pearson, D.J. (1973) Moult of some Palearctic Warblers wintering in Uganda. Bird Study 20, 24–36.

Pimm, S. (1976) Estimation of the duration of bird moult. Condor 78, 550.

Robson, C. (2000) From the Field. Oriental Bird Club Bulletin 32, 66–76.

Round, P.D. (1993) Winter records of the Manchurian Reed-Warbler Acrocephalus (agricola) tangorum from Thailand. Forktail 9, 83– 88.

Round, P.D. (1998) Wildlife, habitats and priorities for conservation in Dong Khanthung proposed NBCA, Champasak Province, Lao PDR. Final Report. Center for Protected Areas and Watershed Management, & Wildlife Conservation Society, Vientiane.

Round, P.D. & Jukmongkol, R. (2001) Recent Reports February– March 2001. Bird Conservation Society of Thailand Bulletin 18, 14–15.

Round, P.D. & Jukmongkol, R. (2002) Recent Reports December 2001–early February 2002. Bird Conservation Society of Thailand Bulletin 19, 14–16.

Seel, D.C. (1976) Moult in five species of Corvidae in Britain. Ibis 118, 491–536

Shibnev, U.B. & Gluschenko, U.N. (1977) [Two new species of reed warblers in the fauna of the Soviet Union]. Abstracts of the VII All-Union Ornithology Conference, part I, Kiev. (In Russian).

Stepanyan, L.S. (1978) Sostav i raspredelenie ptits fauny SSR. Passeriformes. Moscow. (In Russian)

Summers, R.W., Swann, R.L. & Nicholl, M. (1983) The effect of methods on estimates of primary moult duration in the Redshank Tringa totanus. Bird Study 30, 149–156.

Svensson, L. (1992) Identification Guide to European Passerines. 4th edn. Svensson, Stockholm.

Tordoff, A.W. & Eames, J.C. (2001) New additions to the list of birds of Vietnam. Oriental Bird Club Bulletin 33, 37–38.

Underhill, L.G. & Zucchini, W. (1988) A model for avian primary moult. Ibis 130, 358–372.

Vaurie, C. (1959) The birds of the Palearctic fauna: Passeriformes. H.F. & G. Witherby, London.

Williams, M.D. (ed) (2000) Autumn bird migration at Beidahe, China, 1986–1990. Beidahe International Birdwatching Society, Hong Kong.

Williamson, K. (1976) Identification for ringers I: the genera Cettia, Locustella, Acrocephalus and Hippolais. 3rd edn. BTO, Tring.

Kindly submitted by:

Philip D. Round , Department of Biology, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400,
Thailand.

Stephen J. Rumsey, The Wetland Trust, Elms Farm, Pett Lane, Icklesham, Winchelsea, East Sussex TN36 4AH,
UK.