The White-collared Swift (Streptoprocne zonaris) is a large, gregarious, strongly-flying species
found from Mexico and the Greater Antilles to Argentina, and from the Tropical to the Temperate Zone.
Five subspecies were recognized in the standard check-list of Peters (1940), and an additional subspecies
was described by Niethammer in 1953. This study of a series of 225 specimens shows that recognition of
two previously synonymized subspecies was warranted. A large area of the Subtropical Zone of South
America is occupied by a subspecies universally known in the literature as albicincta, but which does not
match the holotype of that subspecies and is hence nameless; it is described herein as new.

Cave swiftlets (Collocalia) from the Mariana, Caroline, and Palau islands build different types
of nests and differ morphologicaIIy from each other and from C. vanikorensis. Populations from the three
respective island groups are here considered specificalIy distinct from C. vanikorensis and each represent
the following separate species: bartschi, inquieta, and pelewensis.

We have estirnated the average mitochondrial DNA nucleotide divergence between Cornmon
and Pallid Swift (Apus apus and A. pallidus) using the restriction fragment technique. These two species
share 63% restriction fragrnents, which corresponds to about 2% sequence divergence. The study of
rnitochondrial DNA sequence divergence may be useful to resolve the unclear phylogenetic relationships
among rnany closely related species of swifts.

The karyotype of Apus affinis presented by Bhunya and Mohanty (1987 showed a remarkable
difference of chromosome number (78) compared to the other three swift species studi ed so far 62, 64 and
62 respectively). Re-examining its chromosome number revealed a discordant result, with only 68
chromosomes. A comparison of the karyotype of A. affinis with that of A. apus shows a great difference in
the number of macro- and microchromosomes, despite a morphological resemblance of macrochromosomes
I to 6. The possible mechanism which can account for the addition of chromosome number of A. affinis is
discussed.

Trans-Anden elevational (above sea leve!) and vertica! (above ground) patterns of distribution of
the resident swifts species in Ecuador were examined. The four types of limits for species distribution along
an elevational gradient proposed by Terborgh (1971) and Terborgh and Weske (1975) were assessed as to
their effects in swift distribution. 1) “Natura! terrninus of the environmenta! gradient” might affect only two
species. 2) “Factors in the physical or biologica! surroundings that change parallel with the gradient” were
possibi!ities for six of lO species. 3) “Competitive exclusion” was not found in the elevational gradient, but,
is likely to operate in partitioning the vertical component. 4) “Vegetational ecotones” did not seem to affect
the aeria! guild.

Homoptera (planthoppers), Diptera (flies), Hymenoptera (social insects), and Isoptera (termites)
were the most numerous prey in 45 food boluses being delivered by parent White-rumped Swiftlets
(Aerodramus spodiopygius chillagoensisi to their chicks inside six Chillagoe caves. The main food items
were planthoppers (47%) and filies (24%), by frequency. The number of insects in each food bolus ranged
from 7 to 587 (x = 149). The average weight of a bolus was 0.33 g (range 0.11 – 0.62 g.). The average
length of ali prey was 3.6 mrn, which is larger than the average length of available prey (2.2 mrn). The
number of.prey “species” ranged from 2 to 83 ex = 40) per bolus. A total of 317 invertebrate “species” were
record ed in food boluses. The White-rumped Swiftlet breeds during the wet season, when insects afe
generally accepted as being more abundant. However, the density of potenti al prey is shown to be
significantly lower than that taken during the breeding season in Fiji.

The diets of Common (Apus apus) and Pallid Swift (A. pallidus) were compared by faecal and
food bolus analysis in a mixed colony in NW Italy. The size of insect-remains increased with age of
nestlings in both species. Size (mm) and mean dry mass of insect prey items was greater in the Common
Swift. There were also differences in the taxonomic composition of prey: the Common Swift took more
aphids in lune, and Heteroptera and Coleoptera in luly, while the Pallid Swift caught more Acalyptera in
lune, and Hymenoptera in luly . Food balls and faecal analysis agreed in their description of swift diets.
A comparison with aerial arthropod abundance, sampled by suction trap, suggested a positive selection of
Hymenoptera and Coleoptera, while Diptera were more frequent in suction trap samples than in the swifts’
diets.

Thirty years ago, a study of Black Swift nesting sites resulted in the establishment of a set of five
nesting site characteristics for this species subsequently confirmed by other researchers. New findings
suggest the modification of one of these and the addition of a new one. These findings are discussed.

Vocalizations of the Common Swifts were studi ed during two consecutive springs in southern
France. I found that three cali types were given by the adults at the nest, and these are described
quantitatively. Significant differences in the acoustic parameters of the calls are highlighted, as well as. a
probable sexual dimorphism. This, together with the precise signification of the different cali types, rernam
however to be critically assessed by playback experiments.

Brood sizes were manipulated to promote different levels of parental effort in the Common
Swift (Apus apus). The two years in which these brood size manipulations were carried out differed with
regard to weather conditions, Data were collected on a visit by visit basis to reveal changes in parental and
chick body rnass, the mass of prey delivered and the estimated mass of parental self-feeding. This provided
a powerful method for testing hypotheses regarding parental investment decision concerned with opti mal
allocation strategies between parents and young and how these can be affected by resource conditions.
When weather conditions were “good” (warmer and sunnier), parents did not have to lower their own selffeeding
to increase the arnount of food delivered to larger broods as they did when conditions were “bad”
(cold and wet), Only in “good” weather conditions did parents suffer no mass loss as a result of increased
parental effort, and incur no increased costs from raising larger broods. In addition, “good” weather
conditions rneant that f1edging mass in larger broods was similar to that in smaller broods, which suggests
that a reduction in the survival chances of fledglings from larger broods only occurred in “bad” weather
conditions. The differenti al allocation responses shown in both years are discussed in terms of parental
strategies to cope with increased brood demands.

Echolocation acuity trials were conducted on Palawan Swiftlets (Aerodramus palawanensis)
under natural conditions in Palawan, Philippine lslands. Detection of 3.2 mm diameter obstacles was
significantly less than for 6.3 mm and 10 mm obstacles. These results are consistent with previous
laboratory trials conducted on other swiftlets. They confirm that although echolocation is used for
orientation in cave nesting and roosting areas it is unlikely to be effective in detecting their typically small
« 5 mm) food iterns.

A series of experiments involving clutch and brood-size manipulation, supplemental feeding
and nest enlargernent were conducted on White-rumped Swiftlets (Aerodramus spodiopygius chillagoensis)
nesting in savannah habitat in Queensland Australia, so that the birds’ reproductive performance might be
compared with that of A.s.assimilis (Tarburton 1987a) which nests in the tropical rainforests of Fiji.
These experiments show that nest-size, predation, synchrony of moult and breeding, and ‘competitive
release’ afe each inadequate to explain why the subspecies in the Queensland savannah has a smaller clutch
than the subspecies in the rainforests of Fiji. While an inadequate food supply prevents Queensland birds
from raising two nestlings at a time it is clear that current interpretations of food limiting theories afe
inadequate to explain why birds of the Queensland savannah produce a smaller clutch than their
conspecifics in rainforests.

L’analyse de 13 années d’observations réalisées sur une colonie de Martinet noir (Apus apus L.)
implantée dans les murs du Museum D’Histoire Naturelle de Nimes (Gard-France), montre que le succès de
reproduction de cene espèce est en relation avec les facteurs climatiques. Nous constatons que certaines
caractéristiques du climat méditerranéeen (par exemple les vents) qui, en immobilisant probablement
l’entomofaune, ralentissent l’accumulation de l’énergie nécessaire à la production d’oeufs, ont un effet sur
l’initialisation, la chronologie et la taille des pontes. De rnèrne, l’action de ces facteurs en période d’élevage
des poussins perturbe l’évolution de leurs courbes pondérales. Si la présence de fortes précipitations reste
rare sous ce type de climat au printemps (l cas en 13 années), nous constatons qu’une varation importante
des températures a des conquénces sur la survie des poussins et celle des adultes reproducteurs. Il ressort de
notre étude que si certains facteurs météorologique semblent plus importants (température, précipitation), ce
sont surtout les vents de secteur nord-nord ouest qui, en agissant de manière plus marquée sur les
disponibilités alimentaires, sont les facteurs les plus determinants quant à la réussite de la reproduction du
Matinet noir.

Survival and fidelity both to the mate and to the nest were studied in a colony of Pallid Swifts in
Piedmont (NW Italy). An overall adult survival rate of 75%, as estimated by the Jolly-Seber and related methods,
was found in the whole colony. Females and males showed no differences in survival rate; the mean
life span as an adult was 3.61 years. Adult breeding birds showed a wide range of strategies: some were
faithful to the nest cavity or to the partner for many years, but a considerable percentage of birds changed
cavity and/or partner year after year. Compared with other European species of swifts, the Pallid Swift
seems to be characterized by slightly lower survival and less intense mate and nest fidelity.

At a small colony in Southern Scotland, long term changes in weather, have been associated
with correponding changes in the date at which the chicks of Common swifts Apus apus, have. reached the
later stages of development. The dates at which the young have reached a suitable stage for n~gmg, have
been significantly influenced by rainfall and temperatures m May and June. Ali of these vanables have
shown regular changes through time. There is also some evidence that July tempe~atures may also have
changed through time and had additional effects. While the strong effects of weather m the early part of the
season are consistent with there being effects of laying date and incubation period on changes in the
breeding schedule, the changes in June and July weather during the later stages of breeding, may have had
effects on chick growth rates which are known to be highly flexible in this species.

We recorded nesting chronology at 15 Vaux’s Swift (Chaetura vauxit nests. Nest building was
observed as early as 3 June. Nestlings were present from 2 July to 4-7 September and were in the nest at
least 27-32 days. Molt of primaries began in early July and probably ended after the swifts left the area.
Weight of 30 adults averaged 18.5 g. One species of feather louse and I species of feather mite were
collected.

Three cases of toe atrophy caused by carpenter ants Camponotus sp. are described in Vaux’s
Swift in northeastem Oregon. We monitored Il Vaux’s Swift (Chaetura vauxi) nests in 1993 in Union,
Umatilla, and Baker Counties in northeastem Oregon, USA.