Breeding biology of the reverse-dichromatic, co-operative parrot Eclectus roratus

Abstract

The breeding biology and social system of the eclectus parrot Eclectus roratus, a species with a unique form of sexual dichromatism (red and blue females, green males) was examined. Our 4-year study at Iron Range National Park on Cape York Peninsula, Australia, showed that females guarded their nest hollows in emergent rainforest trees for up to 9 months each year, often starting well before they laid their first clutches early in the dry season. During this time they rarely left the nest and relied on males for food, both for themselves and their nestlings. Intrasexual competition for scarce hollows and the threat of egg destruction by other eclectus parrots may explain this extreme guarding behaviour. Reproductive success was low. The combined effects of egg and chick loss to conspecifics and predators, brood reduction and flooding of hollows from heavy rain meant that only 18% of eggs and 27% of clutches produced a fledgling. Over 4 years, reproductive success amongst females was strongly skewed: 39% of females were never successful, and only 29% produced more than one fledgling per year. Eclectus parrots were found to breed co-operatively with as many as five males feeding a single female at the nest. Multiple mating, involving one female and four males, was observed once. Despite an even sex ratio at fledging, the adult sex ratio was strongly male-biased, suggesting higher mortality of females. Our data shows that reproductive variance for females is high, and suggests that a shortage of tree hollows and eligible females may enforce co-operative breeding amongst males.

Full text

J. Zool., Lond. (2003) 259,197–208 C
2003 The Zoological Society of London Printed in the United Kingdom DOI:10.1017/S0952836902003138
Breeding biology of the reverse-dichromatic, co-operative parrot
Eclectus roratus
Robert Heinsohn1and Sarah Legge2
1R. Heinsohn Centre for Resource and Environmental Studies, Australian National University, Canberra, A.C.T. 0200 Australia
2School of Botany and Zoology, Australian National University, Canberra, A.C.T. 0200 Australia
(Accepted 29 May 2002 )
Abstract
The breeding biology and social system of the eclectus parrot Eclectus rorat us,aspecies with a unique form of
sexual dichromatism (red and blue females, green males) was examined. Our 4-year study at Iron Range National
Park on Cape York Peninsula, Australia, showed that females guarded their nest hollows in emergent rainforest trees
for up to 9 months each year, often starting well before they laid their first clutches early in the dry season. During
this time they rarely left the nest and relied on males for food, both for themselves and their nestlings. Intrasexual
competition for scarce hollows and the threat of egg destruction by other eclectus parrots may explain this extreme
guarding behaviour. Reproductive success was low. The combined effects of egg and chick loss to conspecifics
and predators, brood reduction and flooding of hollows from heavy rain meant that only 18% of eggs and 27% of
clutches produced a fledgling. Over 4 years, reproductive success amongst females was strongly skewed: 39% of
females were never successful, and only 29% produced more than one fledgling per year. Eclectus parrots were
found to breed co-operatively with as many as five males feeding a single female at the nest. Multiple mating,
involving one female and four males, was observed once. Despite an even sex ratio at fledging, the adult sex ratio
wasstrongly male-biased, suggesting higher mortality of females. Our data shows that reproductive variance for
females is high, and suggests that a shortage of tree hollows and eligible females may enforce co-operative breeding
amongst males.
Key words:Eclectus roratu s,nest hollows, reproductive success, co-operative breeding, dichromatism
INTRODUCTION
Parrots are known for their bright colours and gregarious
behaviour. However, despite roosting communally and
spending at least part of the year in conspicuous
flocks, most are socially monogamous with long-term
pair-bonds. Their apparent monogamy is reflected in
ageneral lack of sexual dimorphism (either colour
or size); sex differences when they do occur are
usually slight (Forshaw & Cooper, 1989; Juniper &
Parr, 1998). Both sexes are involved in parental care,
although the female is often solely responsible for
incubation. Clutch sizes vary from just one egg in
the largest parrots (e.g. some cockatoos, Cacatuinae;
Saunders, 1982) to 10 eggs in smaller species (e.g.
budgerigars Melopsittacus undulates;Wyndam, 1981).
Most parrots nest in pre-existing tree hollows. Although
the conservation implications of these requirements have
been explored in temperate Australia (e.g. Saunders,
Smith & Rowley, 1982; Mawson & Long, 1994; Nelson
&Morris, 1994), very little information is available for
the nesting requirements and breeding biology of tropical
Australasian species (see Marsden, 1992; Marsden &
Jones, 1997 for rare examples).
Eclectus parrots Eclectus roratus are large (500–600 g)
birds found in the Moluccan Islands in Indonesia, the
coastal area and offshore islands of New Guinea, the
Bismarck Archipelago, the Solomon Islands, and Cape
Yo rk Peninsula, Australia. Across this range they are
divided into as many as 12 sub-species; all are found
in lowland and lower montane rainforest (Forshaw &
Cooper, 1989; Higgins, 1999). Eclectus parrots are in
the Psittacidae, and in the absence of clearer taxonomic
affiliation are grouped with the genus Geoffroyus in the
‘unplaced’ tribe Psittaculini (Higgins, 1999). They are
frugivorous and nest in large hollows in tall trees that
emerge above the rainforest (Forshaw & Cooper, 1989).
Eclectus parrots exhibit a variety of traits that set them
apart from other parrots, and indeed from other birds. Most
obvious is their extreme dichromatism: male and female
eclectus parrots are so differently coloured that they
were originally classified as different species (Forshaw &
Cooper, 1989). Females are a spectacular red and blue,
while the larger males are shiny green with a bright

198 R. HEINSOHN AND S. LEGGE
orange beak. The gaudy plumage of females suggests
that sexual selection has operated on females as well as
males, in contrast to the majority of birds and mammals
in which higher reproductive variance for males has led
to their being the sex with elaborate secondary sexual
characters (Andersson, 1994). Reversed size dimorphism
or dichromatism is rare in birds, and is usually associated
with reversed sex roles (see examples in Andersson, 1994).
However, this is not the case in eclectus parrots because,
in typical parrot fashion, the male provides food for the
incubating female, and later for nestlings. In few other
parrot species is the female brighter than the male, and
in all such species the differences are only slight (e.g.
Ruppell’s parrot Poicephalus rueppelli).
Eclectus parrots also differ from most other parrots in
their social system. In contrast to the social monogamy
seen in most parrots, anecdotal reports suggest that several
males feed the breeding female at the nest (Forshaw &
Cooper, 1989). Other parrots with unusual social or mating
systems include polyandrous vasa parrots Coracopsis vasa
(Wilkinson, 1994), communally nesting golden conures
Guaruba guaruba (Oren & Novaes, 1986; Juniper & Parr,
1998); polygamous keas Nestor notabilis (Juniper & Parr,
1998), and lekking kakapos Strigops habroptilus (Best
&Powlesland, 1985). None of these species have marked
sexual dichromatism, and only the keas and kakapos show
strong sexual dimorphism, with males being larger than
females in both species.
Finally, analyses from captive birds show that female
eclectus parrots have remarkable control over the sex of
their young, and often produce long unbroken strings
of one sex (20 males in one case) before switching to
the other (Heinsohn, Legge & Barry, 1997). This is one
of the strongest statistical departures from parity in sex
allocation reported amongst birds (see, Komdeur et al.,
1997; Cockburn, Legge & Double, 2002), but the adaptive
significance of such control remains unknown.
Despite great interest in eclectus parrots from avi-
culturalists and evolutionary biologists alike (Forshaw
&Cooper, 1989; Grafen, 2000), little is known about
their ecology, breeding biology, or social organization (see
Higgins, 1999). The aim of this paper is to provide the first
report of their natural history based on our long-term study
at Iron Range National Park, Cape York, Australia. Here
we confirm that eclectus parrots breed co-operatively,
describe the factors that affect their reproduction, and
provide the foundation for future interpretation of their
unusual social and physical features.
METHODS
Our ongoing study is being conducted on the sub-
species E. r. macgillivrayi at Iron Range National Park on
Cape York Peninsula, in far north Queensland, Australia
(12◦45S, 143◦17E). The national park is located in an
area of lowland rainforest of c.500 km2,which constitutes
about one-half of the entire range of this sub-species. The
rainforest occurs as patches in a complex mosaic that
includes eucalypt woodland and heath. The mean annual
rainfall for the national park is 2780 mm (M. Blackman,
pers. comm.) with most rain falling during a distinct ‘wet’
season from December to April.
This study began in July 1997 and the data presented
covers 4 breeding seasons (1997–2000), with some
additional information from the beginning of the breeding
season in 2001. To find nest trees, we walked along tracks
and creek beds listening for eclectus parrot vocalizations.
Females call loudly at the beginning of the breeding
season, apparently to advertise their presence at the nest
hollow. Fourteen nest trees (with 23 hollows) were located
during July–September 1997, 16 additional nest trees
(with 17 hollows) in 1998, and a further 3 nest trees in
1999 and 2000. Nest trees are used every year, giving a
total of 33 nest trees (with 45 hollows) monitored in this
study. Most such nest trees (n=26) occurred in a block
of c.30km
2of rainforest in the Claudie River drainage
and comprised all nest trees in that area. Our demographic
data are most complete for the eclectus parrots associated
with these trees.
Nest hollows were found in the trunk or a major
branch between 15 and 30 m above the ground. Single-
rope techniques were used to gain access to the nests and
each nest tree was climbed once every 2 weeks to monitor
breeding activity. For each nest visit, records were made
of the presence or absence of the breeding female, the
number of males seen attending her, and the number of
eggs or nestlings in the nest. The lay date of eggs was
established (±3days) by ‘candling’ with a small light.
The progress of the nestlings was monitored and they were
sexed using either down or feather colour. Retrospectively
it was established that female nestlings have darker
grey down than males. Eclectus parrot nestlings develop
directly into their adult plumage colours. From the age of
c.27days, the presence of either green or red pin feathers
on the head indicates their sex. The chicks were banded
on our last visit to each nest when they were c.8weeks
old. On every nest visit, the condition of the nest hollow,
in particular whether it was dry and suitable for breeding
or whether it was damp or flooded due to rain, was also
noted.
Permanent ‘hides’ were built c.20mabove the ground
in trees neighbouring 6 nest trees, and on the ground at 2
nest trees where the hollow was clearly visible. These hides
ranged from 30 to 60 m from the nest hollows and enabled
us to observe the birds at the nest without being detected.
Observations of female presence or absence, group sizes,
and individual identities of colour-banded individuals
were made from these hides. Although eclectus parrots
are difficult to catch, we are able to report the nest-site
fidelity of 8 banded breeding females over multiple years.
It was not possible to distinguish reliably through natural
markings individuals that were not colour-banded. Briefly,
our capture methods entail hoisting mist nets into breaks in
the canopy within 60 m of the nest. When captured, birds
were colour banded with 1 coloured stainless steel band
on each tarsus. These bands were painted using industrial
strength methods (powder-coating) to ensure maximum
duration of the colour.

Breeding biology of Eclectus roratus 199
Tab l e 1. Characteristics of 33 nest trees and 45 nest hollows of Eclectus roratus.Treespecies include: Alstonia scholaris (9), A.
actmophylla (2), Castenospermum australe (5), Ficus a lbipila (5), unidentified (dead) Fi cus sp. (1), Eucalyptus tessellaris (1), Tet ra meles
nudiflora (1), Endospermum myrmecophilum (1), Palaquium galactoxylum (1)
Mean height Mean hole depth Mean entrance width
Genus ntrees nnests (m) (cm) (cm) nfacing sideways
Alstonia 11 14 24.6 ±3.6 SD 85.0 ±30.6 SD 32.9 ±9.6 SD 11/14
Castenospermum 5722.0 ±3.2 SD 87.1 ±28.7 SD 34.3 ±11.3 SD 7/7
Ficus 6924.9 ±3.7 SD 103.0 ±30.8 SD 31.8 ±16.2 SD 8/9
Melaleuca 5619.0 ±1.9 SD 76.7 ±10.3 SD 35.0 ±10.5 SD 2/6
Syzigium 1221.0 (range 6) 70.0 (range 20) 40.0 (range 20) 2/2
Lophostemon 1216.0 (range 4) 90.0 (range 20) 30.0 (range 0) 2/2
Eucalyptus 1215.0 (range 2) 75.0 (range 10) 40.0 (range 0) 1/2
Tetrameles 1120.5 65.0 35.0 1/1
Endospermum 1118.0 80.0 42.0 1/1
Palaquium 1128.0 50.0 45.0 1/1
Total 33 45 22.4 ±4.3 SD 84.4 ±27.2 SD 34.1 ±10.7 SD 36/45
RESULTS
Nest tree and hollow characteristics
All nests were found in hollows in trees that emerged above
the surrounding canopy. Overall, 10 genera of tree were
used by eclectus parrots, but just four genera (Alstonia,
Castenospermum,Ficus,Melaleuca)accounted for 36 out
of 45 hollows in 27 out of 33 nest trees (Table 1). Most nest
trees were completely surrounded by rainforest, except
for the Melaleuca sp. and Eucalyptus tessellaris that were
on the edge of rainforest patches that bordered on open
woodland or swamp. The mean height of nest hollows
above ground was 22.4 m±4.3 SD.Most hollows (35/45)
faced sideways whereas 10/45 opened skywards. Hollows
averaged 84.4 cm ±27.2 SD deep and the mean entrance
width was 34.1 cm ±10.7 SD (Table 1). Two trees had
three simultaneous nests, six trees had two nests, and the
remaining 27 trees had one nest hollow.
The breeding season
Breeding began in the dry season and ended in the wet
season (Fig. 1a,b). All females laid their first clutches
between June and December, and no eggs were laid after
February. There were considerable differences between
years in clutch initiation dates; for example, no female laid
before September in 1997 whereas over half the females
laid their first clutch between June and August in 2001.
Laying dates may have been related to the persistence of
rainfall in the early part of the ‘dry’ season, as females
laid earlier in years when rainfall was lower from May to
June (Fig. 2). However, our sample of 5 years is too small
to confirm this statistically.
Anecdotal reports suggest that breeding usually starts
from June (e.g. Forshaw & Cooper, 1989), but this was
only true in 1 of 5 years. Such reports are most likely an
artefact of the birds’ behaviour. In each year of our study,
females occupied their nest hollows at least from July
onwards, regardless of when they laid eggs. This behaviour
is potentially misleading as females occupy hollows well
DecNovOctSepAugJulJunMayAprMarFebJan
0
5
10
15 1997
1998
1999
2000
2001
(a)
No.of clutches
DecNovOctSepAugJulJunMayAprMarFebJan
0
100
200
300
400
500
600
700
800
Rainfall (mm)
(b)
Fig. 1. (a) Initiation of first clutch by each female Eclectus roratus
vs. month across five breeding seasons. (b) Monthly rainfall records
and mean rainfall for same period.

200 R. HEINSOHN AND S. LEGGE
0.0
Proportion of females
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
100 200 300 400
Total rainfall (May, June, July)
Fig. 2. Proportion of female Eclectus roratus that initiated clutches
by August against dry season rainfall (May–July).
before they breed, so hollows must be inspected to gain
reliable data.
Incubation and nestling periods
Nest trees were not climbed regularly during incubation to
avoid disturbance. The incubation period was 30 days on
five occasions when we were present the day an egg was
laid and just after the same egg hatched. In nine clutches
where precise timing of hatching was known, hatching
intervals between first and second eggs ranged from 1 to
7days(mean =2.8 days ±1.9 SD). In other nests, laying
and hatching dates were estimated from candling eggs
and ageing chicks. The mean nestling period for females
was79days (range =72–86 days) whereas that of males
was86days (range =79–97 days). Our overall sample
of fledged chicks was 51 (24 males, 27 females). Most
clutches were of two eggs, but occasionally a clutch had
one egg (Table 2).
Breeding behaviour
Most females were continually present at their nest
hollows for at least 7 months, and some for as long as
9months. In all 107 female breeding-years, each female
Tab l e 2. Distribution of Eclectus roratus clutch sizes over 4 years
Clutch Clutch Total Total
Year size =2size=1clutches eggs
1997–98 13 4 17 30
1998–99 41 8 49 90
1999–00 31 4 35 66
2000–01 38 7 45 83
Tot al 123 23 146 269
was already present at the nest tree on our first visit to the
nest in July and August each year even if they had not laid
eggs. Females were at the nest hollow on every occasion
the nest was visited (n=5–22 visits per female breeding
season) in 81 out of 107 female breeding-years. Females
in the remaining 26 female breeding-years were present
on between 74% and 95% of visits. Females in 98/107
breeding years were still consistently present at their nest
hollow in February, and females in 36/107 breeding years
were still present in March. None of a limited sample of
females were present in their hollows when checked in
April or May each year (1998, n=12; 1999, n=11;
2000, n=10; 2001, n=8). It was confirmed that females
were present continuously at their hollows throughout the
day by using four all-day watches on two nests in October
1998, and 18 3-h watches on eight nest trees (including
12 nests) from September to December in 1998 and 1999.
During this prolonged nesting period, the females were
fed by at least one attending male. These flew either
to the rim of the hollow or to a nearby branch, and
regurgitated food directly to the begging female. Males
were never observed feeding chicks directly; females
always did this once they had acquired the food from the
males. However, females always stayed at the nest hollow
after chicks fledged either to attempt further reproduction
or to continue guarding the hollow, while the fledglings
and accompanying males went elsewhere. Males seemed
to take on all responsibility for feeding fledglings. Such
observations were rare, but recently fledged juveniles were
seen with males on eight occasions, and never in the
company of their mothers.
Nest site fidelity, female turnover and recruitment
Nest site fidelity data was obtained for eight colour-
banded breeding females over multiple years. Given
various capture dates, it was found that 18 out of 21
females returned to occupy and breed in the same nest
in subsequent years. One additional female returned to
the site to find that her nest tree had fallen over but she
remained in the area. Two females did not use the nest
again; one had died in the hollow from mysterious causes
(see below), and the other disappeared from the study
area and was presumed to have died (Table 3). When five
banded and unbanded females died and two disappeared
during the breeding season, their nests were never taken
over by new females until the following year (Table 4).
It was therefore assumed that hollows with unbanded
females were occupied by the same female throughout the
season. It was also assumed that unbanded females were
faithful to nest sites between years, though we cannot rule
out the possibility that some changes may have occurred
during the non-breeding season.
Of all females, only seven were known to have died
or disappeared over 107 female breeding-years. Two of
these were preyed upon by amethystine pythons Morelia
amethistina in the nest (one 3-m and one 4-m python,
both with large bulge, found in hollow). Two were found
dead but uneaten in the nest, and one was found dead

Breeding biology of Eclectus roratus 201
Tab l e 3. Nest site fidelity for eight colour-banded Eclectus roratus
females. Data were taken until the beginning (July) of the 2000–01
breeding season
Female , year captured Subsequent years Nest site fidelity?
Lofas old hole, 1997 1998–2001 Yes
Lofas new hole, 1998 1999–2001 Yes
Smugglers high hole, 1998 1999–2001 Yes
Mamas, 1998 1999–2001 Yes∗
Syzigium, 1998 1999 Yes
2000 Died
Catfish, 1998 1999–2000 Yes
2001 Died
Smugglers low hole, 1999 2000–2001 Yes
Scrub Itch, 2000 2001 Yes
∗Nest tree fell over in 2001 but female remained at site.
at the base of the nest tree having apparently died from
abroken neck. We suspect that the three females found
dead at or near the nest were victims of female–female
aggression.
Three females that were banded as nestlings occupied
nest hollows and sang to advertise their presence in later
breeding seasons (two in 2000, one in 2001). One was
banded in the first year of the study, and was in her third
year when she bred. She dispersed 6.9 km to occupy a
hollow that faced skywards in a branch of a dead tree to
which we could not gain access, but we know she was
not successful at reproducing. She was not seen again the
following year. Another female was banded in 1998; in her
second year (the 2000 season) she occupied a hollow in
the same tree as another female, c.1.8 km from her natal
hollow. She lost two clutches of eggs, and returned to that
hollow in the 2001 breeding season. The third female was
banded in 1997, and was seen in her fourth year (2001/02
breeding season) occupying a shallow (i.e. inadequate)
hollow in her natal nest tree, although she did not lay a
clutch.
Social group size and composition
Usually, there was only one female associated with each
nest hollow. However, females additional to the number
of nest hollows were seen near resident females without
Tab l e 4. Deaths and disappearance of Eclectus rorat us females, and take-over of nesting hollows
Tak e-over Take-over following
Nest, year Circumstances of female death same season? season?
Funnyslow, 1997 Female dead at base of nest tree No Yes
Funnyslow, 2000 Predation by python in nest No Yes
Scrub Itch, 1997 Female dead in nest No Yes
Syzigium, 1999 Predation by python in nest No No
Tree from hell, 1999 Female dead in nest No Yes
Catfish, 2000 Female disappeared No No
Noisy buggers, 1999 Female disappeared No Yes
No. of nests
123456789
0
2
4
6
8
10
12
14
No. of males
Fig. 3. Number of nests against the mean (hatched bars) and
maximum (shaded bars) number of male Eclectus roratus observed
attending females at those nests.
aggression on more than three occasions during a nesting
period in 23/107 female breeding-years across eight nest
trees. The lack of aggression suggests they may have been
offspring from previous seasons. These additional females
were never seen feeding the breeding female entering the
nest.
The number of males seen around each nest fluctuated
and is summarized in two ways. Figure 3 shows the mean
number of males (rounded to the nearest male) observed
over a minimum of 10 observations from egg laying to
fledging of young, and also the maximum number of males
observed in any one visit over the nesting period. The
mean number of males attending the female ranged from
one (in more than a third of nests) to five. In contrast the
maximum number of males observed ranged from one at
six nests up to nine males at two nests.
Co-operative breeding and multiple mating
Females were fed at the nest by more than one male at
20/34 nests (at 26 nest trees). The greatest number of males
seen feeding females was five; the occasions when more

202 R. HEINSOHN AND S. LEGGE
Tab l e 5. Causes of egg and nestling loss and number of Eclectus roratus fledglings produced over 4 years. Figures are expressed both as
percentage of total eggs and total clutches laid. Percentages of clutches sum to >100% because some processes (e.g. one egg lost to hatch
failure and the second to flooding) applied at the same nest
PartialWhole
Eggs/ Hatch Egg Egg Flooding brood brood
Year clutches Total failure damage disappearance of nest loss loss Fledged
1997 Eggs 30 3 (10%) 4 (12%) 10 (33%) 4 (13%) 3 (10%) 2 (7%) 5 (17%)
Clutches 17 2 (12%) 2 (12%) 5 (17%) 2 (12%) 3 (18%) 1 (6%) 4 (23%)
1998 Eggs 90 11 (12%) 10 (11%) 21 (23%) 9 (10%) 5 (6%) 19 (21%) 15 (17%)
Clutches 49 7 (14%) 6 (12%) 12 (24%) 6 (12%) 5 (10%) 12 (24%) 13 (26%)
1999 Eggs 66 7 (11%) 4 (6%) 20 (30%) 8 (12%) 5 (8%) 13 (20%) 9 (14%)
Clutches 35 7 (20%) 2 (6%) 10 (29%) 5 (14%) 5 (14%) 7 (20%) 7 (20%)
2000 Eggs 83 10 (12%) 1 (1%) 29 (35%) 15 (18%) 2 (2%) 6 (7%) 20 (24%)
Clutches 45 8 (18%) 1 (2%) 16 (36%) 9 (20%) 2 (4%) 4 (9%) 15 (33%)
Tot al Eggs 269 31 (12%) 17 (6%) 80 (30%) 36 (13%) 15 (6%) 40 (15%) 49 (18%)
Clutches 146 24 (16%) 11 (8%) 43 (29%) 22 (15%) 15 (10%) 24 (16%) 39 (27%)
males were observed at the nest did not entail all of the
males feeding the female. We interpret the mean number
of males observed as a conservative approximation of
the stable group size, and the larger numbers of males
observed as temporary. Aggression between males at the
nest was common, and they often displaced each other at
the rim of the hollow when attempting to feed the female.
Groups did not form through natal philopatry of grown
offspring. In 4 years, none of the males banded as nestlings
that fledged successfully (n=24) remained as helpers
in their natal groups. Male attendants were probably
unrelated to the breeding female and were seeking mating
opportunities. We observed multiple mating on one
occasion during the 2000 breeding season. The ‘Mamas’
female, who had fledged two chicks 3 weeks earlier, mated
with four males in quick succession over the space of
c.4min. Each mating occurred as a single episode, and
took between 8 and 15 seconds. There was no apparent
aggression between the males on this occasion. All female
chicks that fledged successfully (n=27) also disappeared
from the natal area, at least for a number of years; three
were observed attempting to breed for the first time (see
above).
Adult sex ratio
Eclectus parrots on Cape York are believed to have a
strongly male-biased adult sex ratio (Forshaw & Cooper,
1989). This is mostly because of the clear predominance in
males seen flying to and from nest trees by most observers,
and the fact that females are rarely seen away from the
nest trees. However, no attempts at quantifying the adult
sex ratio have been made. Using the information above,
arange of possible values from 1.47 to 2.87 males per
female was constructed.
The conservative value (1.47) divides the sum of the
mean numbers of males seen at each of 26 nest trees by
the number of females associated with that tree; additional
females are included if they were seen on at least three
occasions in the course of a breeding season. Females
are treated in this manner as they are less observable;
unlike males, they are not constantly seen returning to
feed the breeding female. The maximum value (2.87)
divides the maximum number of males ever seen at the
nest tree by the maximum number of females. The first
value (1.47) may be too low because it only scores the
number of males seen together at the nest tree, and must
therefore underestimate the number of males attached to
that tree. However, the maximum value is probably too
high as such large numbers of males are only seen at a
nest tree when a female is approaching her fertile period.
Preliminary radio-tracking data confirms that some males
repeatedly visit other nest trees (pers. obs.), so we favour
the lower figure (1.47) as the best conservative estimate of
the adult sex ratio. In either case, males clearly outnumber
females.
Reproductive success
Overall, reproductive success was low, as only 18% of
eggs in 27% of clutches resulted in a fledgling. The causes
of failure at the egg and nestling stages were unexplained
hatching failure, egg damage, egg disappearance, flooding
of the nest hollow, and whole and partial brood loss.
Results for each of these categories, both as percentage of
eggs laid and as percentage of clutches, are given for each
year in Table 5. Table 6 expresses losses resulting from
flooding, and partial and whole brood loss as a percentage
of the total nestlings that hatched.
Twelve per cent of 269 eggs failed to hatch; half of these
did not have a visible embryo and were possibly infertile.
Afurther 6% failed to hatch because of damage (e.g.
breakage) from unknown causes. Other eclectus parrots
probably caused this damage as the eggs were not eaten.
Infertility seemed to be constant across years, whereas egg
damage was most apparent in 1997 and 1998 (Table 5).
Sudden disappearance of clutches accounted for 30% of
all eggs laid. Note that females did not leave damaged eggs
in the nest for more than c.3days, so that it was often not
possible to ascertain whether these eggs had disappeared
because of predation or because of interference from other
eclectus parrots.

Breeding biology of Eclectus roratus 203
Tab l e 6. Causes of Eclectus rorat us nestling loss over 4 years. Figures are expressed as percentage of total chicks hatched and percentage
of total clutches that hatched at least one chick
Nestlings/ Brood Whole brood
Year broods Hatched Flooding reduction loss Fledged
1997 Nestlings 12 2 (17%) 3 (25%) 2 (17%) 5 (42%)
Broods 9 1 (11%) 3 (33%) 1 (11%) 4 (44%)
1998 Nestlings 43 4 (9%) 5 (12%) 19 (44%) 15 (35%)
Broods 34 4 (12%) 5 (15%) 12 (35%) 13 (38%)
1999 Nestlings 31 4 (13%) 5 (16%) 13 (42%) 9 (29%)
Broods 21 2 (10%) 5 (24%) 7 (33%) 7 (33%)
2000 Nestlings 33 5 (15%) 2 (6%) 6 (18%) 20 (61%)
Broods 24 3 (13%) 2 (8%) 4 (17%) 15 (63%)
Tot al Nestlings 119 15 (13%) 15 (13%) 40 (34%) 49 (41%)
Broods 88 10 (11%) 15 (17%) 24 (27%) 39 (44%)
Loss of eggs and chicks as a result of flooding of the
nest hollow from heavy rain accounted for 13% of all
eggs laid. Loss to flooding was highest in the 2000–01
breeding season. Partial brood reduction (loss of just one
No. of breeding females
012 3
0
10
20
30
40
50
60
70 (a)
(b)
Fledglings produced per year
0
Mean breeding success
0–0.5 0.5–1.0 1.0–1.5 1.5–2.0 >2.0
0
2
4
6
8
10
12
Fig. 4. (a) Number of Eclectus roratus fledglings produced in each
of 107 female breeding seasons. (b) Number of females with mean
breeding success (over 3 or 4 years) in six categories.
chick) was relatively rare, accounting for 6% of all eggs
laid (or 13% of nestlings hatched). Whole brood loss,
attributed to either predation or interference, was higher
and accounted for 15% (7–21%) of eggs laid (or 34% of
nestlings) over 4 years.
Over all 107 female breeding-years, 68 failed to produce
any fledglings, 22 produced one fledgling, 13 produced
two, and one produced three fledglings (in two broods) in
1year (Fig. 4a). Of the females that failed in 68 breeding-
years, 38 laid and lost two consecutive clutches, six laid
and lost three clutches, and two laid and lost four clutches.
Of the 22 females that produced one fledgling, six had
already lost one clutch, and of the 14 females that fledged
two or more young, one had already lost one clutch.
There was high skew in reproductive success between
breeding females. Of 28 females monitored for at least
3years, 11 (39%) fledged no young, whereas eight (29%)
averaged over one fledgling per year. Three out of 28
(11%) breeding females averaged over two fledglings per
year (Fig. 4b). Only four females successfully produced a
chick in every year they were monitored; these are referred
to as ‘super-females’.
Only five females attempted a second brood in the
same season after successfully fledging a first brood. Four
of these were the super-females mentioned above. Two
of these females succeeded; the Lofas female fledged
an additional young in 2 consecutive years, and the
Smugglers Bat Hole female succeeded once. Table 7 lists
the success or failure of these five females and, if they
failed to fledge the second brood, the stage at which they
failed.
DISCUSSION
Our monitoring of 146 clutches laid over 107 female
breeding-years at Iron Range National Park yields a
comprehensive picture of the breeding and social biology
of eclectus parrots, an unusual species that has never
been systematically studied in the wild. Our data reveals
specific nest hollow requirements and seasonality of
breeding. Unusual behaviour has also been observed at
the nest including continuous guarding of the hollow
by females and co-operative breeding and multiple

204 R. HEINSOHN AND S. LEGGE
Tab l e 7. Details of 10 attempts to produce a second clutch after
successfully fledging a first clutch for five breeding Eclectus roratu s
females
Female Year attempted Success? Stage failed at
Lofas 1998 No Chick
1999 Yes –
2000 Yes –
Smugglers Bat Hole 1999 No Egg
2000 Yes –
Catfish 1998 No Chick
1999 No Chick
Mamas 1999 No Eggs
2000 No Eggs
Safis 1998 No Chicks
mating by males. Further, there are large and consistent
differences in reproductive success between breeding
females. Here we discuss how these observations provide
an essential foundation for studying the evolutionary
forces responsible for the unusual reverse dichromatism
and sociality seen in eclectus parrots.
Nesting requirements and seasonality
Most parrots nest in tree hollows, and some have specific
nesting requirements which render them vulnerable to
habitat disturbance (e.g. Forshaw & Cooper, 1989;
Mawson & Long, 1994; Nelson & Morris, 1994).
However, few data exist for the hollow requirements of
tropical rainforest parrots (e.g. Marsden & Jones, 1997).
Eclectus parrots used high and deep hollows with small
entrances in a limited number of species of canopy-
emergent trees. These attributes probably provide the best
protection from terrestrial predators. For example, pythons
and monitor lizards Va r anus spp. may have difficulty
climbing large trees when nests are high and there is
no surrounding vegetation to assist in climbing. Height,
good visibility, and a small entrance may also help the
female guard the nest from aerial predators. Given that the
female is brightly coloured and advertises her presence to
other eclectus parrots by calling frequently, there seems
to be little chance for the strategy of concealment. The
small, sideways-facing entrance to the hollow also limits
flooding from heavy rain. However, most hollows flood to
some extent, and one further variable that determines the
quality of a hollow is how quickly it drains. This is more
difficult to measure but probably relates to whether there
are holes in the nest floor, and the porosity (e.g. state of
decay) of the wood.
Most of the hollows were found in four tree genera,
Alstonia sp., Castanospermum sp., Ficus sp., and
Melaleuca sp. Of these, Castanospermum australe,is
endemic to Australia. A strong preference for particular
tree types was also reported for eclectus parrots on the
island of Sumba where 83% of nests were in two species
of Tetrameles (Marsden & Jones, 1997). This preference
probably reflects the tree species that produce the best
hollows. For example, Tetrameles nudiflora is one of the
commonest canopy emergents in Cape York rainforest
butrarely forms hollows. Although it is well established
that breeding densities of many hole-nesting birds can be
limited by a shortage of sites (Newton, 1994), only one
study to date has shown that nest hollow availability can
limit the size of a parrot breeding population (Beissinger
&Bucher, 1992). Data from aerial surveys show that
trees of the required height and species are relatively rare;
there may be as few as 250 nest trees in the 500 km2of
rainforest found in and around Iron Range National Park
(0.5 trees/km2)(S. Legge & R. Heinsohn, pers. obs.). The
extreme hollow guarding behaviour of female eclectus
parrots (discussed below), and documented competition
for nest hollows between eclectus parrots and sulphur-
crested cockatoos Cacatua galerita,also suggests a
shortage of suitable trees (Heinsohn, Murphy & Legge,
in press).
Most anecdotal reports suggest that breeding by
eclectus parrots in most parts of their range is relatively
aseasonal (Juniper & Parr, 1998). In this respect, the
population at Iron Range differs because breeding is
strongly seasonal, with most clutches being laid from
August to December. Although some tree species bear
fruit during the dry season, eclectus parrots do not time
their breeding to coincide with the peak fruiting period
which is usually in the wet season in Australia’s tropical
rainforests (Nix & Kalma, 1972). The seasonality of
breeding in eclectus parrots is similar to that reported for
medium to large parrot species in temperate Australia (e.g.
Saunders, 1982, 1986; Smith & Saunders, 1986; Krebs &
Magrath, 1998) and other seasonal climates such as Pacific
Ocean islands (e.g. red-shining parrot Prosoeia tabuensis;
Rinke, 1989).
The threat of hollow flooding suggests that eclectus
parrots should initiate their clutches as early as possible
in the dry season. They none the less showed considerable
variation in the initiation of breeding between years. The
earliest clutches in 1997 were in September, whereas in
2000 and 2001 many clutches were initiated in June and
July (Fig. 1a). Although the sample size of 5 years is too
low for statistical analysis, timing of egg-laying may be
correlated to rainfall at the start of the dry season, with
females more likely to initiate clutches early when rainfall
from May to July is low (Fig. 2). Two possible reasons for
this pattern are: (1) because their hollows dry out and
become available more quickly when rainfall is low; (2)
that fruit availability depends on the extent of rainfall
during the ‘dry’ season. Longer term data are required to
resolve these issues.
Breeding behaviour
Sex roles, sex ratios, and co-operative breeding
Most parrots are socially monogamous and seem to have
long-term pair bonds (Forshaw & Cooper, 1989). Both
sexes usually contribute to parental care but females
perform all incubation, and are fed by the male during

Breeding biology of Eclectus roratus 205
this period. Usually, females begin to leave the nest
to forage when their nestlings become thermally self-
sufficient (Juniper & Parr, 1998). Female eclectus parrots
are highly unusual as most seem never to leave the vicinity
of the hollow from the time they occupy the breeding site
(July or earlier) until the breeding season ends as late as
March the next year. During this period they call many
times each day, and are aggressive towards other females
and males. When not incubating or brooding, they spend
most of their time sitting at the hollow entrance, with their
bright red heads showing.
Females thus receive all (or most) of their food from
between one and five attending males who regurgitate
it to them at the entrance of the hollow or on a
nearby branch. Nestlings also get all of their food from
the males, but indirectly via their mothers. Females
prevent males from entering the hollow, despite their
repeated attempts to enter or to look inside. Parental
roles change when the young fledge; from this time
on the fledglings stay with the males who feed them
directly. Usually females remain at the hollow for days
or weeks after the young fledge, and continue to visit the
hollow (frequency unknown) throughout the non-breeding
season. The extreme protection of hollows by females
seems to be the principal cause of the departure of eclectus
parrots from the usual parental roles seen in parrots.
Our observations of the number of males attending
females at the nest are based on simple counts of males
present, and how many were observed to be feeding
the females together. Thus, although we can be sure
that multiple males were both present and feeding,
our estimates of group size are probably conservative.
Two sets of data are presented, the maximum number
of males ever observed at a nest during a breeding
episode, and the mean number of males observed over
repeated nest visits. It is important to present the data
in both ways as it suggests that some males visit only
occasionally, perhaps to monitor the female’s reproductive
status, whereas a smaller number seem to be involved in
regular feeding. Preliminary evidence from recent radio-
tracking of adult males (pers. obs.) suggests that males
feed one female primarily but occasionally visit other
nest trees.
None of 24 banded male fledglings remained in their
natal groups. Combined with the male–male aggression
commonly observed at the nest, and our observation of
multiple mating, this suggests that male attendants are
probably unrelated to the breeding female and compete
for mating access. Thus co-operative breeding in eclectus
parrots could take one of two forms. First, it may entail
co-operative polyandry in which males share paternity and
provide care at the nest (e.g. Davies, 1992; Whittingham,
Dunn & Magrath, 1997), although the number of males
present is unusually large for such a system (Hartley &
Davies, 1994). The opportunities for sharing paternity in
one clutch of two eggs are small, but sequential polyandry
is also possible. Second, unrelated males may also help
at the nest as an attempt to gain future reproductive
favour with the female, or female chicks. Such a system is
believed to occur in riflemen Acanthisitta chloris (Sherley,
1989), and pied kingfishers Ceryle rudis,inwhich
unrelated males help without any immediate reproductive
reward (Reyer, 1990).
Naturalists have commented that the adult sex ratio of
eclectus parrots is skewed towards males (e.g. Forshaw &
Cooper, 1989). This study provides the first estimate of the
extent of the skew (between 1.47 and 2.87 males for every
female) although it should be noted that these ratios were
measured at nest trees. It is possible that uncompetitive
males and females that cannot find nest hollows live
elsewhere. However, we have very rarely encountered
female eclectus parrots in our surveys of rainforest
away from nest trees whereas males are commonly seen
foraging in fruiting trees. Consequently we believe that the
sex ratios observed around nest trees are representative of
the entire population.
Since the overall sex ratio at fledging is approximately
even (47% males, n=51, this study; 46% males,
Heinsohn, Legge et al., 1997), the skewed adult sex
ratio suggests that female eclectus parrots suffer higher
mortality than males after fledging. This could result from
their relatively gaudy plumage which renders them more
vulnerable to attack from visually hunting predators such
as raptors. Both rufous owls Ninox rufa,and peregrine
falcons Falco peregrinus,are known predators of female
eclectus parrots (Legge et al., in press). Although the
skewed sex ratio suggests higher mortality for females,
our observations show that breeding females have low
mortality. Since breeding females spend so much time at
their nest hollow, they may be relatively safe because of
the protection the hollow affords, and also because they
are spared potentially dangerous foraging journeys.
Alow number of nest hollows and high female mortality
suggest that breeding females are a limiting resource for
males. This may in turn further necessitate a system
of ‘co-operative’ breeding by males (e.g. Pruett-Jones
&Lewis, 1990) in which males compete for access to
the females. At present, it is not clear whether males
gain access through intra-sexual competition or whether
females choose the best males, or whether both processes
take place (Andersson, 1994). That males are larger than
females (Forshaw & Cooper, 1989), and fight each other
at the nest, certainly suggests a role for intra-sexual
competition.
Fe male turnover and fidelity to nest sites
Our data for colour-banded breeding females suggest
complete nest site fidelity between years. Surviving
banded females returned to the same nest hollow in all 18
subsequent years in which it was possible. In 107 female
breeding-years, five known deaths (two predation, three
mysterious) of breeding females and two disappearances
were documented, all during the breeding season. The
mean annual survival rate for breeding females was
93.7% (90.5%, 100%, 90.9%, and 93.3% from 1997 to
2000). Inter-annual survival in tropical birds is generally
high compared to temperate species (Johnstone et al.,
1997), and has been linked to low rates of territory

206 R. HEINSOHN AND S. LEGGE
turnover in many species (e.g. Greenberg & Dradwohl,
1997; Stutchbury & Morton, 2001). Survival amongst
breeding female eclectus parrots seems especially high,
and stands in contrast to their apparently low survival
between fledging and breeding.
Data for nest site fidelity and female mortality are rare
for large parrots. Some of the best information comes from
studies of white-tailed black cockatoos Calyptorhynchus
funereus.Saunders (1982) found that only one-third of
nesting attempts in this species were made in the same
hollow that had been used previously, and concluded
that hollows were not a limiting resource. Using survival
data from tagged birds, he also concluded that mean
annual survival was 61% and 66% (at two sites) for
adult females and 69% (at both sites) for adult males. In
contrast to white-tailed black cockatoos, the complete nest
site fidelity of breeding female eclectus parrots suggests
that appropriate hollows are limited. Our observations of
intense intra-sexual nest defence by females, and deaths
from apparent fighting, further support this contention.
However, it is intriguing that two hollows remained
unoccupied in the following season, as this suggests that
there were not always ‘floating’ females ready to take over,
possibly because of the differential mortality on females
discussed above.
In our study to date, there were three young females, all
banded as nestlings, that attempted to breed for the first
time. All disappeared for lengthy periods (2–3.5 years),
before reappearing to attempt nesting. Two dispersed (1.8
and 6.9 km) to distant trees, and one returned to its
natal nest tree, so it is not possible to state from this
limited sample whether dispersal is the normal strategy,
as it is for most female birds (Greenwood & Harvey,
1982). However, two of these young females settled in
the same tree as established females, and one was within
80 m of another female. Young females may attempt to
maximize attention from males by settling in an area
already frequented by them. This may explain why two
hollows remained vacant as both were >300 m from
another established hollow. Both young females used
hollows that seemed to be of low quality (e.g. shallow
or facing skywards and therefore prone to flooding). Our
observations also establish a minimum breeding age for
females in the wild of 2 years.
Reproductive success
Eighty-four per cent of clutches in this study had two eggs
and the remainder had one. Parrots lay between one (e.g.
macaws Ara sp.) and 10 eggs (e.g. green-rumped parrotlets
Forpus passerinus;Waltman & Beissinger, 1992), but the
preferred clutch size of two in eclectus parrots is similar to
other large parrots (e.g. some black cockatoos Funereus
spp.; Saunders, 1982). Only 18% of eclectus eggs laid in
146 clutches produced a fledgling, but at least one chick
wasproduced from 26.7% of clutches. Another way to
express overall success is that 51 fledglings were produced,
representing 39 successful attempts from 146 clutches in
107 female breeding-years.
Nest predation in eclectus parrots is unusually high for
hole-nesting birds, which are usually less vulnerable than
open-cup nesters (Martin & Li, 1992). Whereas brood
reduction caused by starvation of nestlings was rare (6% of
eggs), presumed predation of eggs and nestlings accounted
for 45% of all eggs laid. That is, 30% of eggs disappeared
suddenly with no trace of remains, and a further 15%
were lost in a similar manner at the nestling stage. Known
predators were large (3–4 m) amethystine pythons, and
monitor lizards. There was no evidence of any loss of
young to avian predators, but likely candidates are black
butcherbirds Cracticus quoyi,grey goshawks Accipiter
vovaehollandae,and rufous owls. These owls are known
to prey on adult female eclectus parrots (Legge et al., in
press), and may also pose a threat to nestlings.
Breeding success for tropical hole-nesting birds is
usually higher than that reported here for eclectus parrots,
e.g. 78% in New Guinea (Bell, 1982), 67% in the
neotropics (Skutch, 1985), and 66% in a study of buff-
breasted paradise kingfishers Tanysiptera sylvia,also
conducted in the Iron Range area (Legge & Heinsohn,
2001). Breeding success in eclectus parrots is also low
compared to other parrots. Examples of fledging success
ranked from smaller to larger species include 82% of eggs
in green-rumped parrotlets (Waltman & Beissinger, 1992),
50% in crimson rosellas Platycercus elegans (Krebs &
Magrath, 1998), 63% in Hispaniolan parrots Amazona
ventralis (Snyder, Wiley & Kepler, 1987), and 65%
in white-tailed black cockatoos (Saunders, 1982). One
reason for high nest predation in eclectus parrots might be
that they are very noisy around the hollow entrance, and do
not conceal their presence. Males can also be conspicuous
when they return to feed the female, especially when they
squabble with other males near the entrance. Eclectus
parrots may pay a high nest predation cost for their noisy
interactions and for the conspicuous coloration of females.
Thirteen per cent of eggs or young were lost as a result of
flooding of the nest hollow. Most of the flooding occurred
because of sudden storms from October to December,
which varied between years. Other parrot species have
reproductive success that depends on rainfall patterns but
this is manifested in different ways. For some, higher
rainfall leads to better food supplies and an increase in
reproductive success (e.g. western rosellas Platycercus
icterotis,regent parrots Po lytelis anthopeplus,and Port
Lincoln parrots Barnardius zonarius;Long, 1990), and in
others high rainfall leads to hollow damage and exposure
of nestlings (e.g. galahs Cacatua roseicapilla;Rowley,
1990; long-billed corellas, Cacatua tenuirostrus,Smith,
1991). The large losses of young in eclectus parrots
highlight the advantages of having a dry hollow, or of
breeding early enough so that young fledge before the
commencement of heavy rains.
Twelve per cent of eggs failed to hatch for no apparent
reason; about half of these had obvious embryos at
various stages of development. There are three possible
explanations for high rates of hatching failure. First,
hatching failure has been reported to be higher in birds
in lower latitudes (Koenig, 1982; but see Skutch, 1985),
and may result from greater exposure to inappropriate

Breeding biology of Eclectus roratus 207
temperatures (Stoleson & Beissinger, 1999). Second,
inbreeding depression because of high levels of philopatry
may cause hatching failure in some co-operatively
breeding birds (e.g. green woodhoopoes Pheoniculus
purpureus;Ligon & Ligon, 1990). However, this seems
unlikely in eclectus parrots as young disperse from their
natal areas. A third explanation concerns the potential
for decreased heterozygosity and inbreeding depression
resulting from the small size of the Iron Range population,
and the probable bottlenecks experienced in the last glacial
period (Nix & Kalma, 1972). Based on surveys, there are
currently estimated to be 250 breeding females in the
population (S. Legge & R. Heinsohn, pers. obs.). Since
over one-third of these females have never been observed
to breed successfully (see below) the effective population
size is even smaller. These data have strong conservation
implications and will be used for population viability
analyses.
Afurther 6% of eggs were damaged such that
the contents spilled out, or so that the embryo
became desiccated. Other eclectus parrots were probably
responsible for these losses, especially as some eggs had
been shattered with force but with no attempt made
to remove the contents. The real figure could be even
higher, because destroyed eggs were not left in the nest,
and may have been counted as whole clutch or brood
losses in Tables 5 & 6. Although various bird species
are known to destroy the eggs of conspecifics (e.g.
acorn woodpeckers Melanerpes formicivorus;Koenig
et al., 1995), definitive evidence for parrots has only
been presented for the crimson rosella Platycercus
elegans.Inanaviary experiment, Vogels (reported in
Krebs & Magrath, 1998) found that dominant crimson
rosellas destroyed the eggs of subordinate females. Krebs
(1998) also found that egg destruction was common
amongst wild crimson rosellas, and twice observed males
perpetrating the destruction. The aggression shown by
female eclectus parrots towards males suggests that
the latter pose a threat to eggs and nestlings. Egg
destruction may reflect competition between males for
mating access to the female, or if carried out by females
may reflect competition for nest hollows. For example,
female crimson rosellas abandoned their nest boxes if
their clutch had been destroyed (Krebs & Magrath, 1998).
Together with a possible shortage of nesting hollows,
the threat of egg destruction and infanticide suggests an
alternative reason for the extreme protection of hollows
by females. If females can rely on sufficient food from
males, they may be best served by continuous protection
of their eggs and chicks.
Super-females, high reproductive variance for both sexes
Female reproductive success is marked by high skew
between individuals; 39% of individual females observed
over at least three seasons were never successful, and
only 29% produced more than one fledgling per year
on average. Although many females may have failed by
chance, it is also possible that the successful females were
of higher quality (e.g. colour, size, age or experience) or
owned better (i.e. drier, safer) hollows. These two variables
may also be correlated such that higher quality females
gain ownership of better resources (e.g. Bleiweiss, 1985;
Stutchbury & Morton, 2001). However, limited hollow
availability, high mortality of immature females, and high
skew in reproductive success between the small number of
females that do breed has clearly led to high reproductive
variance amongst females. The consequent shortage of
breeding females could in turn be the driving force behind
competition and enforced co-operative breeding amongst
males. This would create strong sexual selection in both
sexes. Further, intense hollow guarding by females has
led to a complete separation of sex roles which, in
combination with strong sexual selection on each sex, may
explain the unique sexual dichromatism seen in eclectus
parrots. The function of colour, combined with molecular
analyses of the mating system, will be the subject of future
research.
Acknowledgements
We thank Peter and Emma Huybers and Mick and
Clare Blackman for their generous support, and Steve
Murphy and Michelle Hall for their help in the field and
comments on the manuscript. Our research was supported
by grants from the Australian Research Council, The
National Geographic Society, and The Winifred Violet
Scott Foundation.
REFERENCES
Andersson, M. (1994). Sexual selection. Princeton, NJ: Princeton
University Press.
Beissinger, S. R. & Bucher, E. H. (1992). Sustainable harvesting of
parrots for conservation. In New World parrots in crisis: 73–115.
Beissinger, S. R. & Snyder, N. F. R. (Eds). Washington, DC:
Smithsonian Institution Press.
Bell, H. L. (1982). A bird community of lowland rain forest in New
Guinea. 2. Seasonality. Emu 82:65–74.
Best, H. & Powlesland, R. (1985). Kakapo. Dunedin: John McIndoe
and New Zealand Wildlife Service.
Bleiweiss, R. (1985). Irridescent polychromatism in a female hum-
mingbird: is it related to feeding strategies? Auk 102:701–713.
Cockburn, A., Legge, S. & Double, M. (2002). Sex allocation in
birds and mammals: disentangling the hypotheses. In Sex ratios:
concepts and research methods: 266–286. Hardy, I. C. W. (Ed.).
Cambridge: Cambridge University Press.
Davies, N. B. (1992). Dunnock behaviour and social evolution.
Oxford, Oxford University Press.
Forshaw, J. M. & Cooper, W. D. (1989). Par ro ts o f the World.
Willoughby: Lansdowne Press.
Grafen, A. (2000). W. D. Hamilton. Guardian 9March: 23.
Greenberg, R. & Dradwohl, J. (1997). Territoriality, adult survival,
and dispersal in the checker-throated antwren in Panama. J. Avian
Biol. 28:103–110.
Greenwood, P. J. & Harvey, P. H. (1982). The natal and breeding
dispersal of birds. Annu. Rev. Ecol. Syst. 13:1–21.
Hartley, I. R. & Davies, N. B. (1994). Limits to cooperative poly-
andry in birds. Proc. R. Soc. Lond. B Biol. Sci. 257:67–73.
Heinsohn, R., Legge, S. & Barry, S. (1997). Extreme bias in sex
allocation in eclectus parrots. Proc. R. Soc. Lond. B Biol. Sci.
264:1325–1329.

208 R. HEINSOHN AND S. LEGGE
Heinsohn, R., Murphy, S. & Legge, S. (In press). Overlap
and competition for nest holes among eclectus parrots, palm
cockatoos and sulphur-crested cockatoos. Aust. J.Zool.
Higgins, P. J. (Ed.) (1999). Handbook of the Australian, New
Zealand, and Antarctic birds. Melbourne: Oxford University
Press.
Johnstone, J. P., Peach, W. J., Gregory, R. D. & White, S. A. (1997).
Survival rates of temperate and tropical passerines: a Trinidadian
perspective. Am. Nat. 150:771–789.
Juniper, T. & Parr, M. (1998). Parrots: a guide to parrots of the
Wo r l d . Sussex: Pica Press.
Koenig, W. D. (1982). Ecological and social factors affecting
hatchability of eggs. Auk 99:526–536.
Koenig, W. D., Mumme, R. L., Stanback, M. T. & Pitelka,
F. A. (1995). Patterns and consequences of egg destruction
among joint-nesting acorn woodpeckers. Anim. Behav. 50:607–
621.
Komdeur,J., Daan, S., Tinbergen, J. & Mateman, C. (1997). Extreme
adaptive modification in sex ratio of the Seychelles warbler’s
eggs. Nature (Lond.) 385:522–525.
Krebs, E. A. & Magrath, R. D. (1998). Breeding biology of crimson
rosellas Platycercus elegans on Black Mountain, Australian
Capital Territory. Aust. J.Zool. 46:119–136.
Legge, S. & Heinsohn, R. (2001). Kingfishers in paradise: the
breeding biology of Tan ysi p terasylvia at the Iron Range National
Park, CapeYork. Aust. J.Zool. 49:85–98.
Legge, S., Heinsohn, R., Blackman, C. & Murphy, S. (In press).
Predation by rufous owls on eclectus parrots and other animals
at Iron Range National Park, Cape York. Corella.
Ligon, J. D. & Ligon, S. H. (1990). Green woodhoopoes: life history
traits and sociality. In Cooperative breeding in birds: 33–65.
Stacey, P. B. & Koenig, W. D. (Eds). Cambridge: Cambridge
University Press.
Long, J. J. (1990). The breeding biology of four species of parrots
in the south of Western Australia. Technical Series No. 6. Perth:
Agriculture Protection Board.
Marsden, S. J. (1992). The distribution, abundance and habitat
preferences of the salmon-crested cockatoo Cacatua moluccensis
on Seram, Indonesia. Bird Conserv. Int. 2:7–14.
Marsden, S. J. & Jones, M. J. (1997). The nesting requirements of
the parrots and hornbill of Sumba, Indonesia. Biol. Conserv. 82:
279–287.
Martin, T. E. & Li, P. (1992). Life history traits of open- vs cavity-
nesting birds. Ecology 73:579–592.
Mawson, P. R. & Long, J. L. (1994). Size and age parameters of nest
trees used by four species of parrot and one species of cockatoo
in south-west Australia. Emu 94:149–155.
Nelson, J. L. & Morris, B. J. (1994). Nesting requirements of
the yellow-tailed black-cockatoo, Calyptorhynchus funereus,
in Eucalyptus regnans forest, and implications for forest
management. Wildl. Res. 267–278.
Newton, I. (1994). The role of nest sites in limiting the numbers
of hole-nesting birds: a review. Biol. Conserv. 70:265–
276.
Nix, H. A. & Kalma, J. D. (1972). Climate as a dominant control in
the biogeography of northern Australia and New Guinea:bridge
and barrier: the natural and cultural history of Torres Strait:
Canberra.
Oren, D. C. & Novaes, F. C. (1986). Observations on the golden
parakeet Aratinga guarouba in northern Brazil. Biol. Conserv.
36:329–337.
Pruett-Jones, S. G. & Lewis, M. J. (1990). Sex ratio and habitat
limitation promote delayed dispersal in superb fairy-wrens.
Nature (Lond.) 348:541–542.
Reyer, H.-U. (1990). Pied kingfishers: ecological causes
and reproductive consequences of cooperative breeding. In
Cooperative breeding in birds:529–557. Stacey, P. B. & Koenig,
W. D. (Eds). Cambridge: Cambridge University Press.
Rinke, D. (1989). The reproductive biology of the red shining parrot
Prosopeia tabuensis on the island of ’Eua, Kingdom of Tonga.
Ibis 131:238–249.
Rowley, I. (1990). Behavioural ecology of the galah Eolophus
roseicapillus in the wheatbelt of Western Australia.Sydney:
Surrey Beatty.
Saunders, D. A. (1982). The breeding behaviour and biology
of the short-billed form of the white-tailed black cockatoo
Calyptorhynchus funereus. Ibis 124:422–455.
Saunders, D. A. (1986). Breeding season, nesting success,
and nestling growth in Carnaby’s cockatoo, Calyptorhynchus
funereus latirostris,over16years at Coomallo Creek, and a
method for assessing the viability of populations in other areas.
Aust. Wildl. Res. 13:261–273.
Saunders, D. A., Smith, G. T. & Rowley, I. (1982). The availability
and dimensions of tree hollows that provide nest sites for
cockatoos (Psittaciformes) in Western Australia. Aust. Wildl. Res.
9:541–556.
Sherley, G. H. (1989). Benefits of courtship-feeding for rifleman
(Acanthisitta chloris)parents. Behaviour 109:309–318.
Skutch, A. F. (1985). Clutch size, nesting success, and predation
on nests of neotropical birds, revisited. Ornithol. Monogr. 36:
575–594.
Smith, G. T. (1991). Breeding ecology of the western long-billed
corella, Cacatua pastinator pastinator. Wildl. Res. 18:91–110.
Smith, G. T. & Saunders, D. A. (1986). Clutch size and productivity
in three sympatric species of cockatoo (Psittaciformes) in
the south-west of Western Australia. Aust. Wildl. Res. 13:
275–285.
Snyder, N. F. R., Wiley, J. W. & Kepler, C. B. (1987). The parrots
of Luquillo: natural history and conservation of the Puerto
Rican parrot. Los Angeles: Western Foundation of Vertebrate
Zoology.
Stoleson, S. H. & Beissinger, S. R. (1999). Hatching asynchrony,
brood reduction, and food limitation in a neotropical parrot. Ecol.
Monogr. 67:131–154.
Stutchbury, B. J. M. & Morton, E. S. (2001). Behavioral ecology of
tropical birds. San Diego: Academic Press.
Waltman, J. R. & Beissinger, S. R. (1992). Breeding biology of
green-rumped parrotlet. Wilson Bull. 104:65–84.
Whittingham, L. A., Dunn, P. O. & Magrath, R. D. (1997).
Relatedness, polyandry, and extra-group paternity in the
cooperatively-breeding white-browed scrubwren. Behav. Ecol.
Sociobiol. 40:261–270
Wilkinson, R. (1994). Vasa parrot’s fascinating breeding behaviour.
Psittascene 6:9.
Wyndam, E. (1981). Breeding and mortality of budgerigars
(Melopsittacus undulatus). Emu 81:240–243.