Revec 0249-7395 2007 Num 62 1 1344
Revec 0249-7395 2007 Num 62 1 1344
Revec 0249-7395 2007 Num 62 1 1344
Nguyên quang Phach, Voisin Jean-François. On an ecological form of the White-nest Swiflet Aerodramus fuciphagus
(Aves, Apodidae) breeding in houses in Viêtnam. In: Revue d'Écologie (La Terre et La Vie), tome 62, n°1, 2007. pp. 49-
57;
doi : https://doi.org/10.3406/revec.2007.1344;
https://www.persee.fr/doc/revec_0249-7395_2007_num_62_1_1344;
Ressources associées :
Aerodramus fuciphagus
In 2003 about 1890-2300 pairs of white-nest swiftlets were discovered breeding in houses in seven towns in southern Viêtnam.
Notable features including their darker rumps, significantly shorter wing length, lower weight, breeding sites exclusively in houses,
extended breeding and moulting periods, separate these birds from Germain’s Swiftlet Aerodramus fuciphagus germani, and put them
near the nominate subspecies
A. fuciphagus fuciphagus of Java. A deliberate introduction of that subspecies seems unlikely. More probably the birds spontaneously
colonised Viêtnam, having been conditioned to breed in houses by farming. For the moment they do not interfere with Germain’s
Swiftlets which breed on offshore islands, but things may change in the future if House Swiftlet numbers grow much. Further studies
are needed to determine the exact taxonomic status of these birds, and for managing its populations sustainably.
Résumé
Sur une forme écologique de la Salangane à nid blanc Aerodramus fuciphagus (Aves, Apodidae) nichant dans les maisons au Viêt-
nam. — Entre 1890 et 2300 couples de Salanganes à nid blanc ont été découverts nichant dans des maisons de sept villes du sud du
Viêt-nam en 2003. Des caractéristiques importantes, comme le croupion plus sombre, la longueur significativement plus courte de
l’aile, le moindre poids, les nids établis exclusivement dans des maisons, des périodes de reproduction et de mue plus longues
distinguent ces oiseaux des Salanganes de Germain Aerodramus fuciphagus germani, et les rapprochent de la sous-espèce
nominative A. fuciphagus fuciphagus de Java. Une introduction délibérée de cette sous-espèce paraît improbable, et il semble plutôt
que ces oiseaux aient colonisé spontanément le Viêt-nam après avoir été conditionnés à nicher dans des maisons par l’élevage. Pour
le moment, on n’a pas observé d’interaction avec les Salanganes de Germain qui nichent sur des îles au large de la côte, mais les
choses pourraient changer dans le futur si les effectifs de «Salanganes des maisons» venaient à s’accroître beaucoup. Des études
plus poussées sont nécessaires pour déterminer le statut taxonomique exact de ces oiseaux, et pour gérer leurs populations de
manière durable.
ON AN ECOLOGICAL FORM
OF THE WHITE-NEST SWIFTLET AERODRAMUS FUCIPHAGUS (AVES, APODIDAE)
BREEDING IN HOUSES IN VIÊTNAM
RÉSUMÉ. — Sur une forme écologique de la Salangane à nid blanc Aerodramus fuciphagus (Aves,
Apodidae) nichant dans les maisons au Viêt-nam. — Entre 1890 et 2300 couples de Salanganes à nid blanc
ont été découverts nichant dans des maisons de sept villes du sud du Viêt-nam en 2003. Des caractéristiques
importantes, comme le croupion plus sombre, la longueur significativement plus courte de l’aile, le moindre
poids, les nids établis exclusivement dans des maisons, des périodes de reproduction et de mue plus longues
distinguent ces oiseaux des Salanganes de Germain Aerodramus fuciphagus germani, et les rapprochent de
la sous-espèce nominative A. fuciphagus fuciphagus de Java. Une introduction délibérée de cette sous-espèce
paraît improbable, et il semble plutôt que ces oiseaux aient colonisé spontanément le Viêt-nam après avoir
été conditionnés à nicher dans des maisons par l’élevage. Pour le moment, on n’a pas observé d’interaction
avec les Salanganes de Germain qui nichent sur des îles au large de la côte, mais les choses pourraient chan-
ger dans le futur si les effectifs de « Salanganes des maisons » venaient à s’accroître beaucoup. Des études
plus poussées sont nécessaires pour déterminer le statut taxonomique exact de ces oiseaux, et pour gérer
leurs populations de manière durable.
SUMMARY. — In 2003 about 1890-2300 pairs of white-nest swiftlets were discovered breeding in houses
in seven towns in southern Viêtnam. Notable features including their darker rumps, significantly shorter wing
length, lower weight, breeding sites exclusively in houses, extended breeding and moulting periods, separate
these birds from Germain’s Swiftlet Aerodramus fuciphagus germani, and put them near the nominate sub-
species A. fuciphagus fuciphagus of Java. A deliberate introduction of that subspecies seems unlikely. More
probably the birds spontaneously colonised Viêtnam, having been conditioned to breed in houses by farming.
For the moment they do not interfere with Germain’s Swiftlets which breed on offshore islands, but things
may change in the future if House Swiftlet numbers grow much. Further studies are needed to determine the
exact taxonomic status of these birds, and for managing its populations sustainably.
1
Khanh Hoa Salangane Nest Company, 38-40 Trân Quy Cap, Nha Trang, Khanh Hoa, Viêtnam
2 Muséum national d’Histoire naturelle, USM 305, CP 51 (Mammifères et Oiseaux), 57 rue Cuvier, 75005 Paris,
France
* corresponding author. E-mail: [email protected]
– 49 –
In 2003, however, during survey operations by one of us (NQP), in the course of a pro-
gramme for assessing the status of swifts and swiftlets in southern Viêtnam, groups of white-
nest swiftlets nesting in houses were discovered in seven mainland urban localities. These
birds differed obviously from local representatives of ssp. germani in several morphological
and biological features. We call them “Viêtnamese House Swiftlets” in the following pages, to
distinguish them from the Viêtnamese White-nest Swiftlet, which we call “Germain’s Swiftlet”
or “ssp. germani”
METHODS
Occupied and potential breeding sites of swiftlets and swifts were surveyed in Viêtnam from Hai Van Pass (16°N)
to Poulo Condor Islands (08°N), in the coastal provinces of Da Nang, Quang Nam, Binh Dinh, Phu Yên, Kahn Hoa,
Ninh Thuan, Binh Thuan, Ho Chi Minh City, and Ba Ria-Vung Tau. Offshore islands, the coast as well as the interior of
the provinces were investigated. In one house in the town of Phan Rang (Ninh Thuan) eleven swiftlets were collected
on 14 October 2003 for anatomical studies. They were killed right after capture and dissected for examination of their
stomach contents, salivary glands and gonads. Their moult stage was evaluated by the BTO method of Ginn & Melville
(1983). In a colony in a house in Nha Trang (Khanh Hoa), the intensity of the swiftlets’ moult was estimated from 17
September 2004 to 12 April 2005 by collecting and counting all moulted feathers on the floor near the nests each tenth
day.
Routine observations on nest building, egg-laying, hatching, etc. were conducted each time a colony was visited
in a house. Supplementary information was also obtained from local people who knew the birds. In order to make
morphological comparisons between subspecies we measured only wing length, since our experience shows that
exposed culmen length is not informative in this species, tarsus length frequently impossible to measure with certainty
on skins, and weight rarely reported on labels.
RESULTS
TABLE I
Numbers of House Swiftlets breeding in Viêtnam in 2004 by localities
TOTAL 10 1890-2300
– 50 –
with closed attics, often replacing the old, lofty ones. Such houses are apparently unfavourable
to the swiftlets. We do not know of any houses built during the past 33 years which has been
colonized by swiftlets.
The House Swift Apus nipalensis (Hogdson, 1837) has an ecology similar to that of swift-
lets and also breeds in houses in Viêtnam. In a few instances we observed apparent competition
for nesting places with Viêtnamese House Swiftlets.
BREEDING
The nest of the Viêtnamese House Swiftlet is a typical “white nest”, entirely made of the
dried secretion of the birds’ sublingual glands. We found eggs or chicks in January, March,
April, September, October, November and December. This corroborates information from
local people, who told us that eggs and chicks could be found during almost every month. The
salivary glands of the birds collected in October were large (0.146 ± 0.05 g) in comparison
with those of ssp. germani at this time, when they are reduced to two thin, thread-like tubes
(Nguyên Quang et al., 2002). In February, however, at a time when ssp. germani is still able to
build nests, the salivary glands of the two forms were of similar size. It seems that Viêtnamese
House Swiftlets build nests over a more extended period than ssp. germani, and may breed all
the year round, or almost so.
Observations made in a Nha Trang colony which had been protected against nest harvest-
ing since 1 September 2004 showed that two breeding bouts occurred between this date and
mid-April (Tab. II). The first bout culminated in early October 2004 when 84.0% of the nests
were occupied by eggs and/or chicks and was completed in late December, when all nests were
empty, although some pairs were busy nest-building at this time. The proportion of nests with
chicks rose abruptly from 23 September to 1 November, and the proportion of empty nests,
taken as an indicator of fledging, rose first slowly until the 1 November, then very rapidly until
TABLE II
Seasonal variation of breeding and moulting intensity in a Khanh Hoa House Swiftlets colony protected since the
1. September 2004
Proportion of
Number of feathers
Date N Occupied nests (%) occupied nests
collected
with chicks (%)
17 Sept. 2004 36 68.5 0 –
23 Sept. 2004 80 78.5 2.5 61
02 Oct. 2004 82 84.0 22.5 47
19 Oct. 2004 82 81.0 * 65.5 –
23 Oct. 2004 82 79.0 * 68.5 25
01 Nov. 2004 82 73.0 * 76.3 24
12 Nov. 2004 74 58.9 * 100.0 25
30 Nov. 2004 74 7.8 * 100.0 55
12 Dec. 2004 74 1.3 * 100.0 47
20 Dec. 2004 74 1.3 * 100.0 53
30 Dec. 2004 74 0 0 59
10 Jan. 2005 94 6.4 0 56
20 Jan. 2005 99 10.1 6.1 35
15 Feb. 2005 100 22.0 6.1 27
28 Feb. 2005 114 50.0 17.9 15
07 Mar. 2005 108 70.1 20.1 15
15 Mar. 2005 115 73.0 23.5 9
26 Mar. 2005 124 74.5 39.0 74
04 Apr. 2005 124 67.8 * 49.2 39
12 Apr. 2005 124 66.1 * 52.4 47
Note. — N: number of nests under observation (these values differ from the ones in Tab. VII, which are the total
numbers of nests in the colony); *: no new egg; -: no data. Empty nests afford for deserted nests, nests where breeding
was unsuccessful and nests left by fledglings; occupied nests for nests with eggs and/or chicks.
– 51 –
the 30 December, when all chicks had left their nests. The second bout culminated at the end
of March 2005, but nest-occupation was still high by mid-April 2005.
When all nests in a colony are collected simultaneously, the birds start to rebuild about
15 days later and complete the replacement nests within 40-45 days. They thus breed more or
less synchronically. However, not all colonies are harvested completely at the same time. If a
few nests escape harvesting, pairs at all stages of breeding may be found at the same time in
the same house.
MOULT
Examination of ten of the eleven collected Viêtnamese House Swiftlets (one was excluded
because its plumage was damaged upon capture) revealed that five had fresh wing quills, and
the other five were in the process of moulting their 5th or 6th primary, which meant that they
already had moulted primaries 1 to 4 or 5. These five last birds had fresh secondaries. Four
birds of that sample of ten were moulting either their 1st or 5th rectrices. Viêtnamese House
Swiftlets have thus an extended moulting period, covering a large part of the year and largely
overlapping with the breeding season, contrary to ssp. germani in Viêtnam, in which it is sea-
sonal (Nguyên Quang et al., 2002).
Moulting birds were found in the Nha Trang colony quoted above throughout the period
of observation, but the intensity of moulting, expressed in quantity of moulted feathers gath-
ered at every visit, was not constant (Tab. II). Moulting reached a peak in each breeding bout
just after most pairs had completed egg-laying. It also reached a secondary peak just after the
fledglings had left their nests. By contrast, moult was minimum when the birds were building
their nests or feeding their chicks (Tab. II), two highly energy-demanding processes.
TABLE III
Measurements and weights of 11 Viêtnamese House Swiftlets breeding in the town of Phan Rang (Viêtnam)
– 52 –
THE FOOD OF ADULTS
The stomachs of the eleven collected specimens of Viêtnamese House Swiftlets contained
together 652 food items (Tab. IV). Because many of them were partly digested, and swiftlets do
not deliver food to their chicks from their stomachs, but from a sublingual pouch, those food
items may be considered as being adult, and not chick, food. About half of those items could
not be identified because of digestion damage.
As far as we can infer from our small sample (Tab. IV) the food of the Viêtnamese House
Swiftlet is predominantly based on Diptera, followed by Hymenoptera and Coleoptera. On
the contrary, the food of adult Germain’s Swiftlets, ssp. germani consists predominantly of
Hymenoptera, followed by Isoptera and Diptera, and nestlings of ssp. germani are fed on about
20% Diptera, their staple food being Homoptera (Tab. V) (Nguyên Quang et al., 2002; Voisin
et al., 2005). Thus, it is probable that competition between the two forms is small as far as food
is concerned. Competition for food between the Viêtnamese House Swiftlet and the Black-nest
Swiftlet A. maximus is probably also of small importance, as the latter species eats over 80%
large Hymenoptera in Viêtnam (Nguyen Quang, 1996) (Tab. VI). The diets of the three swiftlet
taxa which breed in Viêtnam and their feeding competition deserve more studies based on
larger samples.
TABLE IV
Occurence of prey items in the food of 11 Viêtnamese House Swiftlets breeding in the town of Phan Rang (Viêtnam)
TABLE V
Prey items of Germain’s Swiftlet Aerodramus fuciphagus germani in Viêtnam (after Nguyên Quang et al., 2002)
– 53 –
TABLE VI
Occurrence of prey items in stomach contents of 15 adult Black-nest Swiftlets
Aerodramus maximus in Khanh Hoa (Nguyên Quang, 1996)
DISCUSSION
– 54 –
TABLE VII
Wing length, weight and plumage coloration of the subspecies of the White-nest Swiftlet Aerodramus fuciphagus
(Thunberg, 1812)
Note. — Differences checked with Student’s t-test. ns: mean not differing significantly from that of Viêtnamese house
swiftlets; * and **: mean differing significantly from that of Viêtnamese house swiftlets (** p < 0,01; * p < 0,1).
the systematic status of which will be the subject of another study. No other measurement of
White-nest Swiftlets from the Anamba Islands seem to have been published. According to
Oberholser’s description, amechanus has a darker rump than germani, but as the type speci-
mens were moulting, the colour difference could be due to plumage abrasion (see below).
Their dark grey-brown upperparts with paler rumps distinguish Viêtnamese House Swift-
lets from subspecies micans (Stresemann, 1914) from the Lesser Sunda Islands, which is dis-
tinctly paler and greyer, and from vestitus, from Sumatra and Borneo, which is very dark with
a concolorous rump. Their measurements tell them also apart from inexpectatus (Hume, 1873)
of the Andaman and Nicobar Islands, and dammermani (Rensch, 1931), known only by the
type from Flores and of doubtful validity (Nguyên Quang et al., 2002).
Comparison with museum specimens shows that, while Viêtnamese House Swiftlets
do not differ significantly in wing length from the form cf. amechanus from North Borneo
(Tabs. III & VII), they have paler, browner backs, contrasting more with their rumps, and paler
underparts. Viêtnamese House Swiftlets do not differ significantly in wing length from the
nominate form fuciphagus (Thunberg, 1812) from Java and neighbouring islands (Tabs. II &
III). Our specimens are very similar to these in under- and upperparts colouration, and in rump
colour, but their upperparts are clearly, even if slightly, darker than in the Javan form. In these
– 55 –
characters, the Viêtnamese House Swiftlet is not distinguishable from Aerodramus fuciphagus
fuciphagus. The slightly dark upperparts of our specimens remain unexplained. It may be an
effect of sampling, or due to the fact that the original population from which those birds come
from is itself slightly darker than the average Javan birds.
One interesting point is that Viêtnamese House Swiftlets’ fledglings in Viêtnam have paler
rumps than adults. As adult plumage is attained within one year, it is unlikely that it occurs as
a consequence of moulting. Close observation of the rump of adult A. fuciphagus from several
localities shows that the exposed part of each feather is more or less brownish grey in the mid-
dle, with pale or whitish margins which can be more or less worn. This is especially true of
feathers covering the midline of the rump. The overall appearance of the rump depends on the
intensity and extent of the grey colour (Cranbrook, 1984; Nguyên Quang et al., 2002). In these
conditions, adult rump colouration may be acquired through abrasion of the feather margins,
a mechanism well-known in birds. In Germain’s Swiftlet, the grey colour in the middle of
the rump feathers is distinctly less dark than in ssp. fuciphagus, and the whitish margins are
broader. As a result, the rump of the bird appears all the time pale, even if there is some varia-
tion. This point deserves further study.
ORIGINS
A deliberate introduction of House Swiftlets in Viêtnam in the 70s seems unlikely, because
this technique practised at that time involved fostering the eggs or hatchlings under Collo-
calia esculenta, a species which does not exist in Viêtnam. Moreover, swiftlets resembling A.
fuciphagus fuciphagus have spontaneously colonized houses in Sumatra and along the west
coast of the Malay Peninsula in recent years (Cranbrook, com. pers.). A last reason is that,
when House Swiftlets turned up in Viêtnam, prices for nests were still low (Lau & Melville
1994), and thus breeding these birds was much less rewarding than now. The settling down
of House Swiftlets in Viêtnam may then be the result of a natural range extension from the
south by White-nest Swiftlets which have been conditioned to houses for breeding after having
hatched and brought up in “farms”, that is, in houses.
ACKNOWLEDGMENTS
We wish to thank warmly the Earl of Cranbrook for having considerably improved the manuscript and given
us much good advice. We thank also Mr. A. Bens, Dr. J. Cuisin, Mr E. Gros, Dr. P.-Y. Henry, Dr. Chr. Jouanin,
Mrs S. Marchand, Dr. V. Nicolas and Dr. C. Voisin who compared with highly concording results our House Swiftlet
specimens with specimens of Aerodramus f. fuciphagus and A. f. germani. Our thanks go also to Dr. R. Prys-Jones for
the loan of specimens from the Natural History Museum at Tring (Herts.), and to Mrs Anne Préviato who made pictures
of the specimens.
– 56 –
REFERENCES
– 57 –