Previous Brood Research Articles

Daughters on request: about helpers and egg sexes in the Seychelles warbler.

The Seychelles warbler (Acrocephalus sechellensis) was an endangered endemic of the Seychelles islands where, until 1988, the entire population of ca. 320 birds was restricted to the one island of Cousin Island (29 ha). Additional breeding populations were successfully established on the islands of Aride (68 ha, 1988) and Cousine (26 ha, 1990) and now with the existence of ca. 2000 warblers on three islands the conservation status of the warbler has improved from endangered to vulnerable. Emigration from the island is extremely rare, so birds that disappeared were known to have died. Almost every bird on Cousin Island has been individually colour ringed and monitored throughout all breeding attempts during a 17-year period (1985-2002; total ca. 2400 birds). These birds were also blood sampled for molecular parentage and sex analyses. Therefore the lifetime reproductive success of many birds is known. Although warblers can breed independently in their first year, some individuals remain in their natal territory as subordinates, and often help by providing nourishment to non-descendant offspring. The frequency of 'helping' is affected by habitat saturation and variation in territory quality (insect prey availability). The long-term benefits of helping are higher for daughters than for sons, and it is therefore no wonder that most helpers are daughters from previous broods. Furthermore, on low-quality territories breeding pairs raising sons gain higher fitness benefits than by raising daughters, and vice versa on high-quality territories. Female breeders adaptively modify the sex of their single-egg clutches according to territory quality: male eggs on low quality and female eggs on high quality. The Seychelles warbler is a beautiful example of behavioural and life-history adaptations to restricted circumstances.

Adaptations and maladaptations to island living in the Seychelles Warbler

The Seychelles Warbler (Acrocephalus sechellensis) was an endangered endemic of the Seychelles islands where, until 1988, the entire population of ca. 320 birds was restricted to the one island of Cousin Island (29 ha). Although warblers can breed independently in their first year, some individuals remain in their natal territory as subordinates, and often help by providing nourishment to non-descendent offspring. The frequency of helping is affected by habitat saturation, variation in territory quality (insect prey availability), and the genetic relatedness between the helper and the offspring. Helping results in indirect benefits from enhancing the reproductive success of close relatives, and direct benefits as improved parental skills and the acquisition of parentage. The overall helping benefits are higher for daughters than for sons, and it is therefore no wonder that most helpers are daughters from previous broods. Furthermore, on low-quality territories breeding pairs raising sons gain higher fitness benefits than by raising daughters, and vice versa on high-quality territories. Female breeders adaptively modify the sex of their single-egg clutches according to territory quality: male eggs on low quality and female eggs on high quality. However, despite the saturated nature of the Cousin population, the possibility of obtaining higher reproductive success on new nearby island, and a well developed flight apparatus, inter-island dispersal by Seychelles Warblers is extremely rare. The Seychelles Warbler is a beautiful example of behavioural and life history adaptations and maladaptations to restricted circumstances.

The reproductive cycle of thelytokous colonies of Cerapachys biroi Forel (Formicidae, Cerapachyinae)

The reproductive cycle of thelytokous colonies of Cerapachys biroi was investigated in laboratory conditions. It included several stages, the duration of which was assessed by direct observation on experimental nests set up by fission of natural, queenless, colonies. The results show that two phases of adult activity alternated while cohorts of brood developed synchronously. During the statary phase, virgin workers laid diploid eggs simultaneously but no foraging activity was observed until the following phase which began after the callow workers of the previous brood cohort emerged and the eggs hatched. During this foraging phase, workers explored intensively for food while larvae matured synchronously. Biometrical measures allowed to determine that three larval instars occurred in this species. While larvae started pupation, workers stopped to forage and the next statary phase took place, during which a new cohort of eggs was laid and pupae matured. This repeated phasic reproductive cycle where adult activity and development of the brood are synchronised reminds that of true army ants (subfamily Ecitoninae), which brings new arguments in favour of a filiation between both groups.

The breeding biology of the Mangrove Gerygone, Gerygone laevigaster, in the Darwin region, with notes on brood parasitism by the Little Bronze-cuckoo, Chrysococcyx minutillus

Gerygones comprise 19 species of small (mostly 6–7 g) foliage-gleaning insectivores distributed from south-east Asia to New Zealand. I collected breeding data on the Mangrove Gerygone, Gerygone laevigaster, from several populations in and around Darwin, Northern Territory. The egg-laying season extended from March to October, coinciding with the late wet season and entire dry season, with peaks in April and October (n = 45). All populations showed repeat nesting after failure; one colour-banded pair made four attempts over one season. The species is apparently multi-brooded, one pair attempting a late brood after two previous successful broods. Most complete clutches (74%) were of two eggs, and the remainder, three. Most nests (84%) were built in Grey Mangroves, Avicennia marina, on the edge of saltflats, and 91% were in the top half of the nest plant. Incubation and nestling periods were 16–17 and 14–17 days, respectively. Of 31 nests whose fate was known, only three (10%) produced fledglings. Five broods (16%) were parasitised by Little Bronze-Cuckoos, Chrysococcyx minutillus; the incubation period for one cuckoo egg was about 15 days, while two nestling periods were estimated to be 18 and 24 days. Several nests were inundated by high spring tides, but predation was probably the major cause of nest failure: seven (23%) nests exhibited holes, six (19%) were destroyed, while eight (26%) showed no damage. The long, mainly dry-season, breeding of this species may be an adaption to greater insect availability at this time of the year, or avoidance of wet-season predators (such as reptiles), or both.

Double-Brooding in Red-Necked Grebes

-Four attempts of double-brooding (rearing two broods in one year) in the Red-necked Grebe (Podiceps grisegena) in SE Poland were recorded out of 163 breeding pairs studied over the period 1993-1999. Two pairs successfully reared two broods during the same nesting season. Time appears to be an important constraint on the incidence of double-brooding. The pairs attempting double-brooding were among the earliest breeders and the second clutches were started early, relative to the hatching time of the previous brood. All pairs attempting a second brood nested on fry ponds with high fish availability. Double-brooded pairs fed the first-brood young more frequently before the onset and after termination of incubation of the second clutch than did single-brooded pairs during the corresponding period of brood rearing. However, in comparison with single-breeders, parents attempting double-broods were aggressive to the first-brood young earlier and fed them less frequently during the incubation of the subsequent clutch. Apparently only early breeding on territories with abundant food allows grebe parents to overcome conflicting pressures resulting from overlapping parental investment in successive reproductive attempts. Received 8 March 2000, accepted 5 October 2000.

Patterns of brood production in the grass shrimp Palaemonetes pugio (Decapoda: Caridea)

Summary The pattern of brood production was analyzed in the grass shrimp Palaemonetes pugio. Determination of patterns of brood production as successive (spawning after each molt) or alternate (spawning after every other molt) was made by observations on individual females paired with males in the laboratory over a four-month period. A pattern of successive parturial (spawning) molts was the most common pattern. However, in a majority of these successive parturials, the ovary of the female was not mature at hatching of embryos from a previous brood. There was a delay of 10–11 days, during which ovarian maturation took place, between embryo hatching and the next (parturial) molt. This pattern of “slow successive” parturials does not result in the truly continuous brood production which occurs in many caridean species, i.e., spawning of a new brood within 1–2 days after hatching of a previous brood. However, there was a relatively low percentage of females in which the ovary was mature at embryo hatching. These latter females did have a parturial molt after only 2 days (“fast successive” parturial). In the “alternate” pattern, more frequent later in the 4-month observation period, hatching of embryos was followed by a non-parturial (without spawning) molt within 2 days, followed by an intermolt period of 10 days during which ovarian maturation occurred prior to the next (parturial) molt. The interval between spawns was similar in the “slow successive” and “alternate” parturial molt patterns (22–23 days) compared to a much shorter one (14–15 days) in the “fast successive” parturials. Examination of reproductive females from field samples indicated that the spawning pattern observed in the laboratory was that which occurs in natural populations in southern Louisiana. Although brood production in P. pugio is not truly continuous, the relatively short interspawn intervals observed still allow production of several broods during the breeding season.

Female Reproductive Cycle and Sexual Conflict over Precopulatory Mate‐guarding in Thermosphaeroma (Crustacea, Isopoda)

In species with time‐limited opportunities for insemination, precopulatory mate‐guarding is expected to coevolve with the duration of female reproductive cycles. Despite this adaptation to female characteristics, it may also be advantageous for males to adjust the duration of guarding with respect to sex ratio because the benefits of guarding are dependent on the availability of females. If female fitness is reduced because of guarding, male guarding behavior leads to intersexual conflict. We studied these aspects of male mate‐guarding behavior in two closely related, thermal‐spring isopods (Thermosphaeroma). First, guarding duration showed species specificity which was related to the duration of reproductive cycle; cycle length for females and duration of guarding by males in T. milleri were twice as long as in T. thermophilum. Second, males in both species adjusted their guarding duration with sex ratio, guarding longer when a competing male was present. Third, in T. thermophilum, ovarian development began immediately after the birth of the previous brood and continued through guarding, sexual molt and post‐molt periods until oviposition, whereas in T. milleri, ovarian development was largely postponed until the post‐molt period. Because guarding during ovary provisioning periods may be costly for females, we tested the existence of intersexual conflict over guarding duration in T. thermophilum. We compared the duration of guarding of control pairs with those of pairs in which either male guarding ability or female ability to resist guarding was reduced experimentally. Guarding durations for manipulated and control males were equal, but manipulated females were guarded longer, suggesting that conflict exists and that females can effectively shorten guarding duration by their behavior. Moreover, we suggest that selection in the context of intersexual conflict may play an important role in the evolution of delayed oviposition and sperm‐storage organs in mate‐guarding crustaceans.

Survival costs of chick rearing in black‐legged kittiwakes

1. We tested for costs of chick rearing in the black‐legged kittiwake Rissa tridactyla (Linnaeus) by removing entire clutches from 149 of 405 randomly selected nests, in which one or both mates was colour‐banded. After the manipulation, we monitored adult nest attendance and body condition at unmanipulated and manipulated nests, and measured the survival and fecundity of these adults the following year. 2. Late in the chick‐rearing period, adults from unmanipulated nests (i.e. with chicks) went on significantly longer foraging trips, and were significantly lighter for their size, than adults from manipulated nests (i.e. without chicks). 3. Adults from unmanipulated nests also survived to the following nesting season at a significantly lower rate than those from the manipulated nests (0·898 vs. 0·953), suggesting that attempting to raise chicks can reduce life expectancy by 55%. 4. There was a tendency for adults from nests that were unmanipulated in year one to have lower reproductive success in year two, primarily because of reduced fledging success, and a higher incidence of non‐breeding. 5. These findings suggest that mass loss in kittiwakes during chick rearing may not be adaptive. Raising chicks can lead to reproductive costs, and the causal mechanism appears to be a reduction in body condition. 6. We compare our results with previous brood (or clutch) size manipulation experiments that have measured adult body condition, survival and/or future fecundity. Although the empirical evidence suggests that long‐lived species are more likely to experience survival costs than short‐lived species, we believe the opposite may be true. We suggest that shifting the experimental protocol of cost of reproduction studies from brood enlargements (an approach taken in most prior studies) to brood reductions will provide more accurate quantifications of naturally occurring costs. 7. The cost of reproduction is one mechanism proposed to explain the reduced survival rates reported for kittiwake populations in the North Atlantic relative to those in the North Pacific ocean. Oceanographic data, however, suggest that lower food availability may limit survival of kittiwakes in the North Atlantic where a deeper mixed layer and reduced primary production combine to make conditions less favourable for this seabird during the winter months.

Number of reproductive cycles of Varroa jacobsoni in honey-bee (Apis mellifera) colonies

Very little data exists concerning the number of reproductive cycles performed by individual Varroa mites. To understand the population dynamics of the Varroa mite it is necessary to know the number of fertile female offspring each Varroa female produces during her lifetime. The lifetime reproduction capacity of the mite consists of the mean number of fertile female offspring produced during each reproductive cycle multiplied by the mean number of cell passages. This paper describes an experimental design to estimate the number of reproductive cycles where mites are transferred to new mite-free colonies for reproduction in sealed brood cells. The data presented suggests that the mean number of reproductive cycles performed by the individual female mite is larger than previously accepted. Under optimal conditions, the mean number of reproductive cycles by Varroa females is probably greater than 1.5 but less than 2. Furthermore, the results show that the reproductive success of Varroa females going into cells to reproduce is not influenced by previous brood cycles.

A within-clutch trade-off between egg production and rearing in birds

Here we compare the chick-rearing capacity of common terns Sterna hirundo that were given an additional chick at hatching (incurring rearing costs but no production costs), an additional egg at the start of incubation (incurring incubation and rearing costs) and made to lay an additional egg (incurring laying, incubation and rearing costs). The brood enlargements were within the natural range of brood sizes (i.e. two increased to three). Only those pairs that were given a free chick reared a significantly larger brood compared with unmanipulated controls. The results show for the first time a within-clutch trade-off between producing and rearing extra young. Although there was no effect on egg or hatchling size as a result of the increased egg production, parents incurring the full costs of producing extra young subsequently showed reduced chick provisioning, growth and survival. These results cast doubts on the interpretation of previous brood enlargement experiments as providing empirical evidence that observed clutch sizes are often less than the Lack clutch size, as such experiments have failed to include production costs.

The influence of egg survivorship on the subsequent nest fidelity of female bicolour damselfish, Stegastes partitus

Abstract. In many animal species, females lay eggs within male territories. Research on female choice in such species has shown that females often assess male or territory quality indirectly on the basis of male characters that are correlated with male or territory quality. Researchers have given much less attention to the possibility that females can assess male or territory quality directly based on the survival of previous broods. This study demonstrates that in the bicolour damselfish, a species with exclusive male parental care of eggs, females use the survival of previously-laid eggs to avoid nests that have a low probability of offspring survival. Females mate repeatedly during a reproductive cycle and although they generally show high nest fidelity, this fidelity is broken when a nest loses an entire brood to nocturnally feeding brittlestars. Experimental removal of broods and addition of brittlestar odours to nests demonstrate that brittlestar odours are sufficient to cause females to avoid nests. Females may also respond to visual cues associated with brood loss. In addition, clutches laid in nests after nocturnal complete-brood losses are much less likely to survive than clutches laid in nests that have not suffered nocturnal complete-brood losses. Therefore, by avoiding nests from which brittlestars have recently removed eggs, females increase the probability of future egg survival.

Breeding of Collared DovesStreptopelia decaoctoin rural Oxfordshire, England

Breeding of Collared Doves was studied for 3 years in rural Oxfordshire. They laid eggs from February to October, mostly between April and mid-August. In rural Britain, breeding starts and finishes about one month later than at similar latitudes in eastern Europe although the populations have been separated for less than 50 years. The reasons for this are not clear. Collared Doves laid on average 3.8 clutches per year. Of completed clutches 89% were of 2 eggs, the rest of 1. Most pairs began a new clutch while still feeding fledged young, a few while the previous brood were still in the nest. Mean hatching success was 59%, fledging success 69%, and overall breeding success 41%. Success varied with season, peak production coming from eggs laid in May to July. Losses of eggs and chicks were mainly from predation, especially early in the season. Mean annual productivity was 3.1 young per pair. By feeding on super-abundant, predictable and persistent supplies of commercial crops, Collared Doves have the potential for high productivity. Predation by humans has increased substantially in recent years and this has probably brought Collared Dove populations in Britain to near stability after logarithmic population growth following colonization about 30 years ago.

The Breeding Biology of the House Sparrow Passer domesticus in Israel

The breeding biology of the House Sparrow was studied in the Mediterranean coastal plain of Israel. Egg laying occurred from the end of March until mid July. The mean clutch size was 5.05 eggs with a modal clutch size of 5, whereas the most productive clutch size was 6 eggs. On average each pair laid 2.3 clutches per year from which a total of 5.75 young fledged. Mean egg weight increased with laying sequence within the clutch. Incubation length was negatively correlated with ambient temperature. Average breeding success was 70.0%. Nestling growth rate was negatively correlated with brood size, but both the time the nestlings spent in the nest and nestling feeding rate were positively correlated with brood size. The time interval between broods increased by on average 1.6 d for each additional chick raised in the previous brood. These results are compared with those obtained for other populations of House Sparrows.

Satiation restores brood care motivation in the female mouthbrooder Pseudocrenilabrus multicolor (Cichlidae)

The maternal mouthbrooder Pseudocrenilabrus mulitcolor takes up her eggs during spawning and carries them continuously in her mouth for 10–11 days at 27°C. Then the young—by that time able to swim—are released from the mouth. The duration of mouthbrooding is controlled by an internal timing program in the mother [12]. In non-first-spawners brood care motivation (=readiness to take up brood into the mouth and keep it there) disappears within a few hours after brood removal. At the same time hunger—which is reduced to half of the normal level when brood care motivation is present—increases to the normal level [15]. In females without brood care motivation, satiation by food intake or by a water solution of 15 mg/l diethylpropion (an anorectic amphetamine-derivative), restores brood care motivation during a sensitive period of approximately 11 days after spawning. The point of release of adopted brood from the mouth depends on the duration of previous brood deprivation, but not on the state of brood care motivation nor on the hunger state during brood deprivation. The results suggest that above a certain level hunger inhibits brood care motivation. By contrast, the hunger state seems not to influence internal timing of mouthbrooding.

Brood Deprivation and the Control of Mouthbrooding in Pseudocrenilabrus multicolor (Cichlidae)

and Summary The female of the mouthbrooding cichlid Pseudocrenilabrus multicolor from East Africa takes up her eggs during spawning and carries them continuously in her mouth for approximately 10 days at 27 °C. Then the young — now able to swim — are released from her mouth. A former study showed that the point of release is under the control of an internal timing program in the mother which can be only slightly modified by age-specific stimuli from the brood. The present investigation is concerned with how brood deprivation influences mouthbrooding in first spawners. If the brood is removed from the mother's mouth for longer periods of time (several days), the internal timing program can be split into two components that control different behavioural parameters: 1 Brood care motivation, i.e. the readiness to take up brood into the mouth and keep it there but not eat it, is maintained up to 10 to 11 days after spawning (at 27 °C), regardless of the duration of previous brood deprivation. 2 In contrast, the point of release of brood from the mouth is retarded as a linear function of the duration of previous brood deprivation. Additionally, brood older than 10 days at the point of release is released one day earlier than younger brood. The mechanisms of these different time-measuring systems are discussed with the aid of a model which assumes internal hour-glass timers. The adaptive significance of the mechanisms remains unclear.

BIOLOGY OF THE LITTLE BITTERN IN THE SOUTHWESTERN CAPE

Summary Langley, C. H. 1983. Biology of the Little Bittern in the Southwestern Cape. Ostrich 54:83-94. The breeding biology of the Little Bittern Ixobrychus minutus payesii was studied at the Rondevlei Bird Sanctuary, Southwestern Cape, South Africa, from 1975 to 1979. Adult behaviour before and during nesting is described. Adults are diurnal. Bills of breeding birds flush red during courtship, copulation and nest relief. Breeding occurred from June to February, adults occupying the same nest area each season. Nest and nest site are described. Mean clutch size was 33 eggs and mean measurements were 34,6 x 26,6 mm. Average incubation period was 19 days. Chicks left the nest permanently at 14–16 days and flew at 27 days. Development and behaviour of young are described. Hatching success was 62,3% and nestling success 56,6%. Additional broods occurred on average 16,8 days after the previous brood had flown.

Intraseasonal Reproductive Costs for the House Sparrow (Passer domesticus)

Abstract House Sparrows (Passer domesticus) near Calgary, Alberta begin breeding in early spring and continue through to late summer. High productivity from previous broods is negatively correlated with fledgling production from second and third broods. Although fat reserves may limit the ability of females to raise young, there is no concomitant drop in clutch size or in the probability of renesting. Pairs that fledge many young in a year space fledgling production evenly over the breeding season but are most productive in mid-season. The interval between fledging and the initiation of the next clutch increases with the number fledged. This delay, an indication of the physiological strain involved in rearing young, is greater for later broods and for females nesting in trees. Measures of reproductive effort (clutch size, number fledged, length of the nestling period) vary seasonally but give no indication of peaking for last broods. Thus, reproductive effort is not adjusted to parallel changes in the probability of surviving to the next breeding attempt.

The Breeding and Feeding Characteristics of Darwin's Finches on Isla Genovesa, Galapagos

The breeding and feeding of four species of finches were studied on Isla Genovesa, Galapagos from January to May 1978. Food supply and finch diets were studied in the nonbreeding season (November) as well. The species in order of decreasing body size were Geospiza magnirostris, G. conirostris, G. difficilis, and Certhidea olivacea. Initiation of breeding was not a simple function of body size; the medium—sized cactus finch, G. conirostris, started breeding first, followed by G. difficilis and G. magnirostris in that order. Eggs of G. conirostris were first laid within °1 wk of the first rains. Certhidea laid a maximum of three clutches, G. magnirostris and G. conirostris laid up to four, and one pair of G. difficilis laid five. Similarities among all four species were found in incubation periods (10—14 d, mode 12 d), nestling periods (11—17 d, modes 13—15 d), and uniform success of clutches of different sizes. Similarities among the three Geospiza species were found in the interval between successive clutches (°30 d); the modal size of clutches (four eggs); seasonal variation in clutch sizes, in which average size was largest among second clutches and smallest among fourth clutches; egg and nestling production; and the average interval between fledging from one nest and egg laying by the parents in the next nest. This average interval was 1—4 d, but some pairs of all three species started laying 5—6 d before the nestlings of the previous brood had fledged. The species differed in the number of fledglings produced per pair; from the first three clutches G. difficilis produced 9.5 ± 0.5 per pair, G. conirostris produced 7.9 ± 0.7, G. magnirostris produced 5.1 ± 0.9, and Certhidea produced 3.2 ± 1.5. The relatively low success of G. magnirostris was attributed to predation of nestlings by owls (Asio flammeus), which occurred most frequently within 1 wk of expected fledging. The low success of Certhidea is attributed to disturbance at the nest by G. difficilis and mockingbirds (Nesomimus parvulus), and possible predation by mockingbirds. Compared with finch species in a climatically similar region of the Ecuadorian mainland, Darwin's Finches have large clutches, a much longer nestling period, higher nesting and breeding success, but similar, and low, incidences of egg infertility. Certhidea fed nestlings on an exclusively arthropod diet, whereas the three Geospiza species fed nestlings on a mixed diet of arthropods, fruits, and seeds. All Geospiza species fed size selectively on spiders in proportion to their average body and beak size. Diets of the three Geospiza species were largely overlapping throughout the breeding season. Differences reflected the differences in beak sizes. Seeds of Bursera graveolens, which are difficult to crack, were absent from nestling diets of (the smallest) G. difficilis, and most frequent in nestling diets of (the largest) G. magnirostris. Only G. conirostris fed nestlings on parts of Opuntia helleri (cactus). Diets of the three species diverged markedly in the nonbreeding season, at which time they clearly reflected the adaptations of beak sizes and shapes. The results are discussed in terms of the strong seasonality in food production and availability, coupled with interannual variation in amount, onset, and pattern of rainfall, and hence food production. It is concluded that adaptations for food gathering in the nonbreeding season, when food is scarce, have indirect influences upon reproductive traits, resulting in small but significant demographic differences between species.

Social Organization and Behaviour of the Grey-Crowned BabblerPomatostomus Temporalis

SUMMARYKing, B. R. 1980. Social organization and behaviour of the Grey-crowned Babbler Pomatostomus temporalis. Emu 80: 59–76.The social system of wild and captive Grey-crowned Babblers Pomatostomus temporalis was studied in south-eastern Queensland. Three classes of age were distinguished on colour of iris and a fourth by difference in plumage. Adult birds could be sexed by length of bill. The ratio of males to females in a population did not differ significantly from unity. The species occurred in sedentary territorial groups, each of which usually contained a single breeding pair and a number of non-breeding birds. Non-breeding members were of all ages and mostly the offspring of previous broods of the breeding pair. Groups moved as closely knit units and much behaviour was directed towards cohesion of the group. Patterns of behaviour and calls are described. Groups were strongly territorial and spent much time in disputes with neighbouring groups. In general, size of home range did not directly correlate with size of group. All members of a group co-operate in building and care of the young but only the breeding female incubates and broods. The degree of participation of an individual in visiting nests and feeding young usually corresponds with its age and status within the group.

The Influence of Nests on the Social Behaviour of Males in Manorina Melanocephala, a Communally Breeding Honeyeater

Manorina melanocephala is a communally breeding honeyeater in which several males feed the offspring of a single female. A free-living population of colour-banded birds was observed from 1971 to 1974 in southern Queensland.Males visiting nests were frequently attacked by males previously dominant at that place. Immatures from a female’s previous broods in the same year were driven away from the nest by adult males. Sometimes males displayed towards the eggs in the nest and often drove off the incubating female before such display. Females almost never visited the nests of other females and only rarely fed their fledged offspring. Females appeared to learn to leave the nests at the approach of a male and thus incubation could possibly be considerably interrupted when many males visit. Every nest watched had a complement of visiting males, whose number increased as the nest advanced. At least twenty-two males were known to visit one successful nest. Feeding rates of nestlings by males reaching fifty-five per hour; of fledgelings eighty per h our. Females’ contributions were additional to these. Males did not restrict their activities to a single nest but often had a much more active role at one. The largest recorded complement (banded males only) attending successive nests of a single female in the same breeding period numbered twenty-four. Spaces in which females were active showed no clustering nor were they confined to the area occupied by a coterie of males. Of nests built by fourteen banded females, an average of 5-2 were visited by thirty-three banded males during systematic watches. The mean number of females visited was 3'8. Males ranking highest in visiting any female also visited more nests and more females than did lower ranking males. No male ranked highest with more than one female. Often females were visited by males from more than one coterie but then always from an adjacent one. When a female nested within the area of a coterie, most visiting males originated there. When a female nested near the common boundary of one or more coteries, males from each visited. Visiting males tended to show a linear relation between rank in visiting activity and distance of their space of activity from a nest. The social system, in general, is complex and unlike any previously described for birds. Although observations of early nests in the breeding period suggest synchrony between close females, analysis did not support widespread synchrony of reproductive activity within a colony.