Behaviour and breeding biology of the cooperatively breeding Grey-backed Fiscal Shrike Lanius excubitorius in Kenya

Grey-backed Fiscal Shrikes Lanius excubitorius were studied over a 2j year period near Lake Naivasha, Kenya. Grey-backs are cooperative breeders, with group sizes ranging from two to II. Only one pair breeds per group, with all other group members aiding in the rearing of young. The study population ranged from 64 to 79 individuals that occurred in from 13 to 16 groups. Non-breeding helpers made up to 66% of the population, with male helpers being more numerous overall than females. The annual survival rate was 65%, with no differences detected between the survival of males and females, or of breeders and helpers. Only male helpers were observed to acquire breeding status within the natal territory. Some female helpers acquired breeding positions in territories adjacent to their natal territories. Group territorial displays occurred throughout the year but were most pronounced prior to breeding during rainy periods. Reproductive success was very low, with only 14.5% of the recorded breeding attempts leading to fledged young. Large groups (four or more birds) had greater reproductive success than small groups, but many factors other than, or in addition to, group size may have influenced this pattern. The breeding male contributed the most food to the incubating female and to the nestlings. Male helpers and the breeding female contributed more to nestlings than did female helpers. Observations on the post-fledging period indicate that socialization and establishment of dominance may be of importance in group-living species living in a restricted ecological and social setting.     

THE PRESENT AND PAST DISTRIBUTION OF THE BALD IBIS IN THE PROVINCE OF THE CAPE OF GOOD HOPE

Tarboton, W. R. 1981. Cooperative breeding and group territoriality in the Black Tit. Ostrich 52:216-225.

In a small, colour-ringed population of Black Tits Parus niger in central Transvaal, 11 of 19 observed breeding units comprised pairs with one to three helper-males. These pairs and groups defended permanent territories, the size of which correlated with the size of the group. There were significantly more territorial disputes during winter when less food was available than in summer. Breeding occurred in summer and the female alone built the nest, incubated the eggs and brooded the young while they were small. During this time she was fed by the alpha male and helper males, although before egg-laying the alpha male prevented helpers from courtship-feeding her. On average, unassisted pairs reared 0,88 young/season whereas pairs with helpers reared 1,55 young/season. However the feeding rate of nestlings of pairs with helpers was not higher than that of unassisted pairs and the number of young reared per group did not correlate with the number of helpers within the group.

The helper system in Black Tits was associated with a skewed sex-ratio (1,7:1 males: females) in the adult population and the data are consistent with the “hopeful reproductive” hypothesis for cooperative breeding.     

Reproductive success and helper effects in the cooperatively breeding grey-crowned babbler

Cooperative breeding, where some individuals help to raise offspring that are not their own, is a relatively rare social system in birds. We studied the breeding biology of a declining cooperative breeder, the grey-crowned babbler Pomatostomus temporalis , with the aim of isolating the social factors that affect its reproductive success. Most breeding pairs were assisted by philopatric offspring, although pairs could breed successfully without helpers. Females laid up to four clutches (usually three eggs per clutch) per season. Male (but not female) helpers increased the number of young fledged from individual nests and the likelihood of re-nesting, resulting in higher seasonal fledgling production. Helper effects on brood size and fledgling production were greater in the second year of the study, which was also characterized by higher nest failure. This suggests that helpers enhance reproduction more in poor conditions. Our study demonstrates the interacting effects of social and ecological factors on reproductive success, and that retention of offspring is not always beneficial for the breeders in cooperative species.     

The Social System of a Communal Breeder,the Bay-winged Cowbird Molothrus badius

Molothrus badius (bay-winged cowbird), an icterine blackbird with cooperative breeding, shares behavioural and ecological characteristics with other communal nesters: it is sedentary, has a high annual survival rate (76.2%) and a strong nest-site tenacity (mean breeding dispersal of 41.9 and 89.4 m for males and females). Behavioural data, including collective agonistic displays, suggest group territoriality. Before egg hatching most breeders occurred as single pairs showing territorial behaviour (82% of nests), and nesting was usually solitary (distances to nearest nests of 25–103 m). Most breeders were apparently monogamous, with a 2.5% incidence of extrapair copulations in the territory during clutch formation. During the nestling stage one to four helpers occurred at 95% of M. badius nests. Most helpers were 1–2 years old, but older breeding adults (mostly males) that failed to rear their own offspring helped at the end of the season. The number of helpers increased (up to 4) with nestling age. Helpers were also recruited during the postfledging period, and group size reached up to 10 adults at this stage. Helpers mobbed predators and brood parasites, and provided 35% of the nestling food. Provisioning rate was positively and significantly correlated with number of helpers, although age of nestlings was the best predictor of overall food delivery rate. The helping system was almost obligate and productivity comparisons between nests with/without helpers are not possible. Data suggest that helpers increased the breeding success per nest. The correlation between the provisioning rates of parents and helpers was negative but non-significant. In 18% of nests 3 to 4 individuals were present before the nestling period, including cases of apparently polyandrous trios and one case of joint nesting by two pairs. Within Brown 's (1987) categories of social organization M. badius is mainly group territorial with plural nesting. Habitat requirements of M. badius are wide and nest sites do not appear to limit breeding. Kinship plays a role in the social system, as 9 of 12 helpers marked as nestlings helped their parents.     

Sex-bias and timing of natal dispersal in cooperatively breeding Puff-throated Bulbuls Alophoixus pallidus

While natal dispersal can have a significant impact on population dynamics, it is typically difficult to quantify. We investigated timing of natal dispersal of the cooperatively breeding Puff-throated Bulbul Alophoixus pallidus in a tropical evergreen forest by modelling the probability of staying in or dispersing from their natal territory whilst taking into account the effects of sex, group size, and the presence of helper(s). Birds did not disperse until the beginning of and during the breeding season following the hatching year. Dispersal was strongly female-biased both in frequency and distance: most females (95%) dispersed away from their natal territories, and of those relocated, traversed 2–7 territories. In contrast, 50% of males remained in the natal territory as helpers in their second year, while relocated dispersing males crossed 1–2 territories. Natal dispersal was not influenced by either group size or the presence of helpers. Males that fledged earlier in the breeding season exhibited higher rates of philopatry than the males that fledged later, but no correlation between fledging date and philopatry was observed in females. The probability of staying in the natal territory during the second year was 0.58 ± 0.14 SE and 0.05 ± 0.04 for males and females, respectively. These findings may add to our understanding of how natal dispersal can reflect social patterns and kin structure in cooperative breeding species from a little-studied tropical forest region.     

Southern Lapwing Vanellus chilensis cooperative helpers at nests are older siblings

Cooperative breeding is rare in shorebirds, and when found it is thought to be due to polygamous mating (cooperative polygamy). Here we describe the social structure of cooperatively breeding groups in Southern Lapwing Vanellus chilensis and test the prediction that offspring sex ratio is skewed towards the sex that helps. The social groups consisted of a breeding pair with one or two young (mostly males, 68%) from the previous breeding season, and offspring sex ratio was not skewed. Cooperative breeding in the Southern Lapwing is not the consequence of cooperative polygamy, but rather groups were formed by a mated pair and some of their offspring from previous nests as helpers.     

Genetic monogamy in single-male groups of acorn woodpeckers, Melanerpes formicivorus

Acorn woodpeckers have one of the most complex social systems of any bird species. Breeding units range in size from monogamous pairs to groups of 15 birds that include multiple breeding males and females as well as nonreproductive helpers-at-the-nest. Groups form when young remain at their natal nest to help their parents breed or when single-sex coalitions of siblings disperse to fill a reproductive vacancy on another territory. Plural breeding and helping behaviour are thought to be favoured through indirect fitness benefits for individuals that would otherwise be unable to breed due to a shortage of reproductive vacancies on territories with acorn stores. We report the results of multi-locus DNA fingerprinting of 51 offspring from 18 nests of 16 socially monogamous pairs of acorn woodpeckers. If socially monogamous females mate outside the pair-bond, indirect fitness benefits for cobreeders and helpers will be significantly reduced. Monogamous pairs accounted for all but one of the 51 offspring we tested; the single exception was apparently sired by the putative father, but the putative mother was excluded from maternity. Our results indicate that individuals remaining on their natal territories as helpers are generally the genetic offspring of the pair they help. They also suggest that single-sex coalitions offspring dispersing together from nests of socially monogamous pairs will be full-siblings.     

Broodcare helpers in the cichlid fish Lamprologus brichardi: Their costs and benefits

‘Helping’ in birds and mammals involves seemingly altruistic behaviour. In the cichlid fish Lamprologus birchardi helpers are usually young of former broods staying in their parents' territories and participating in all kinds of parental duties (Broodcare, territory maintenance and defence). The discovery of helpers in fish offered the chance of attempting an extensive analysis of potential costs and benefits influencing the evolution of helpers in a vertebrate. Three factors proved to be of major importance in the cost-benefit analysis of helping as opposed to leaving for family-independent nonreproductive aggregations. Due to investment and to their rank within a family's hierarchy, helpers grow at a slower rate than non-helpers. This cost is compensated for by (i) a lower mortality risk to helpers caused by their access to a defended shelter and by protection afforded by bigger family members, and (ii) a positive contribution by helpers to the future reproductive success of their parents: females with helpers produce bigger clutches and consequently more free-swimming fry (=siblings). Other variables, such as the helpers' influence on the relative breeding success of their parents, broodcare experience through helping, the chances of territory take-over, parasitism of parents' reproduction and cannibalism are of minor importance. Similar social organizations in other fish are discussed with respect to their ecology and are compared with cooperatively breeding birds and mammals.     

Shared parentage and incest avoidance in the cooperatively breeding acorn woodpecker

Social groups of acorn woodpeckers (Melanerpes formicivorus) range in size from unaided pairs to 15 adults. Behavioural indicators of mate guarding, assumed incest avoidance and observations of egg-laying indicate that social organization ranges from monogamous pairs to groups with up to seven male and three female putative cobreeders plus up to 10 nonbreeding helpers. In addition, groups occasionally lack a putative breeder throughout the breeding season. Here we report results from multilocus DNA fingerprinting of 372 nestlings from 123 nests in groups with putative cobreeders of one or both sexes. No extra-group fertilizations were found. Putative cobreeding males within social groups shared paternity. However, the most reproductively successful male was, on average, almost three times as successful as the next most successful and additional males only occasionally sired offspring. In contrast, cobreeding females shared parentage equally. Helpers never bred incestuously when their opposite-sex parent (or another relative, such as their uncle) held breeding status in the group. However, during breeding male vacancies, 14 nestlings were produced when helper males bred incestuously with their mother. Both male and female helpers usually became successful cobreeders with their same-sex parent following replacement of the opposite-sex breeder(s) by unrelated individuals.     

Geographical variation in patterns of parentage and relatedness in the co‐operatively breeding Ground Tit Parus humilis

The Ground Tit is a large parid species endemic to the Tibetan Plateau. Here we describe its genetic breeding system based on 2 years of fieldwork on a population from Damshung, Tibet. Genetic relatedness and parentage were analysed using 16 microsatellite markers and sex was determined with a marker on the Z and W chromosomes. We established that 16 of 75 families (21%) were assisted by one or occasionally two male helpers, which in most cases were young from a previous brood of one or both of the breeding adults. The helpers typically stayed with their families throughout the breeding season. Helpers never obtained any paternity in their own families, but one helper was identified as the sire in the only instance of extra‐group (and extra‐pair) parentage detected. Thus, the level of extra‐group/extra‐pair parentage appears to be very low in this Ground Tit population, and sharing of reproduction within family groups is apparently absent. Our results contrast with the findings from another Ground Tit population in Gansu, further northeast in China, with respect to both the number and sex of helpers and the division of parentage within and among family groups. In Gansu, helpers regularly produced offspring and both extra‐pair and extra‐group paternity as well as maternity (egg dumping) was common. Differences in family structure, philopatry, territoriality and potential inclusive fitness benefits can probably explain this contrast, and are likely to reflect the relative costs and benefits of co‐operative breeding.     

Group structure,within-group conflict and reproductive tactics in cooperatively breeding Galápagos mockingbirds,Nesomimus parvulus

Reproductive conflict within groups can be an important feature of cooperative breeding systems, especially when more than one individual of a sex breeds within a social group. Relationships between group structure, dominance, within-group conflict and reproductive tactics of cooperatively breeding Galápagos mockingbirds were examined on Isla Genovesa. Territorial groups of 2–24 adults included up to three breeding females, with 42% of the groups containing more than one (plural groups); females in most plural groups nested separately. Territory size increased with group size, but the area available per pair in plural groups was smaller than in singular groups (groups with only one breeding pair). Most pairings were monogamous, and males usually outnumbered females; high-ranking males obtained mates more frequently than subordinate males. In 3 relatively dry years, but not in a wet El Niño year, subordinate pairs in plural groups fledged fewer young than dominant pairs or pairs breeding in singular groups. Interference by dominant breeders, often leading to abandonment of nests by subordinate pairs, appears to account for these differences: through nest disruption in drier years, dominant individuals may reduce the cost of sharing their territories and increase the chances of recruiting helpers. Dominant males in plural groups may also father young through extra-pair copulations with subordinate females. Despite costs imposed by within-group conflict, subordinate breeders have higher long-term reproductive success than birds that defer breeding. Plural group structure is maintained because unpredictable climatic variation favours opportunistic breeding by subordinates.     

Relative reproductive success of female moorhens using conditional strategies of brood parasitism and parental care

In a population of moorhens (Gallinula chloropus), at least27% of netting females laid one or more eggs in a neighbor'snest Females laid parasitically under three conditions: 56%of parasitic eggs were from nesting females that preceded layinga dutch in their own nest by a parasitic laying bout, 19% werefrom females whose nests were depredated before clutch completionand that laid the following egg parasiticaDy, and 25% were froma small number of females without territories, "non-nesting"parasites, that each laid a series of parasitic eggs. Clutchsizes varied greatly between females, but nesting females eachlaid a consistent clutch size both within and between seasonsfor a given mate and territory. Nesting females that employeda dual strategy of brood parasitism and parental care producedextra eggs that they laid in the nests of neighbors before layinga dutch in their own nests. Two out of ten females whose dutchesI experimentally removed during the laying period were successfullyinduced to lay their next egg in the nest of a neighbor. Nestingfemales that laid parasitically selected their hosts opportunisticallyfrom among the nests dosest to their territories. An experimentin which parasitic eggs were removed and hosts left to rearonly their own young showed that parasites did not choose hoststhat were better parents than pairs with contemporary neststhat were not parasitized. Females that only laid parasiticaDywithin a given season timed their parasitic laying bouts poorlyand achieved no reproductive success. Parasitic young rarelyfledged, and the mean seasonal reproductive success of nestingbrood parasites did not differ from that of nonparasitic females.However, the variance in reproductive success of nesting broodparasites was significantly higher than that of nonparasiticfemales.     

FACTORS AFFECTING BREEDING OF RAZORBILLS ALCA TORDA ON SKOKHOLM

A study of the breeding biology of the Razorbill was carried out on Skokholm (South Wales) during 1971-73. Birds ringed or colour ringed before the study began provided additional information upon the effects of age on breeding. Mean laying date was delayed in 1972, compared with 1971; the effect is attributed chiefly to stormy weather which upset colony attendance. Eggs were also smaller in 1972. A seasonal decline in egg size (volume) was noted in all three years, attributed mainly to the later laying of young birds. Egg size increased with age, at least up to the fifteenth year. Eggs lost totalled 30% of those laid; 73% of this total was due to predation by Herring Gulls and of Jackdaws. Most losses (45%) occurred during the first 10 days after laying. Of lost eggs, 25% were replaced, usually 14 days after the loss of the original; only eggs laid and lost early in the season could be replaced. Only 7% of the chicks which hatched failed to fledge. Most (62.5%) chick losses occurred in the first week of nestling life, when chick weight was related to egg size. Afterwards, both growth rate and fledging weight were independent of egg size. The chicks fledging early in the season were heavier than later chicks. Failure to fledge was mainly due to a breakdown in behaviour between parent and young, rather than to predation. Breeding success was highest for birds breeding early in the season, most of which were older, more experienced breeders. These laid early enough to replace an egg if it was lost; they produced large eggs, and their chicks were therefore both heavier than average during the critical first 7–10 days of life, and fledged at a high weight. Thus experience accumulated with age, and the ability to lay early in the season are important for successful breeding in the Razorbill.     

Pair territoriality in the longnose filefish,oxymonacanthus longirostris

The longnose filefish,Oxymonacanthus longirostris, usually lives in heterosexual pairs, the male and female swimming together and sharing the same territory. Pair territoriality in the species was examined in detail in relation to sexual differences in territorial defense activities. Rigorous pair territoriality was maintained only during the breeding season, although pairs used their home ranges exclusively to a certain extent, during the non-breeding season. The frequency of aggression against other conspecific pairs in the breeding season was higher than in the non-breeding season. Agonistic interactions appear to be over both mates and food resources, the strict pair territoriality in the breeding season possibly being due to mutual mate guarding. In intraspecific aggressive interactions, males usually led their partner females when attacking intruders. The feeding frequency of males was much lower than that of females in the breeding season. Mate removal experiments indicated that females could not defend their original territories solitarily and their feeding frequency decreased. Conversely, males could defend territories solitarily without a decrease in feeding frequency. These results suggest that males contribute most to the defense of the pair territory, with females benefiting from territorial pair-swimming with their partner males.     

Egg investment in response to helper presence in cooperatively breeding Tibetan ground tits

Life‐history theory predicts a trade‐off between current and future reproduction to maximize lifetime fitness. In cooperatively breeding species, where offspring care is shared between breeders and helpers, helper presence may influence the female breeders’ egg investment, and consequently, survival and future reproductive success. For example, female breeders may reduce egg investment in response to helper presence if this reduction is compensated by helpers during provisioning. Alternatively, female breeders may increase egg investment in response to helper presence if helpers allow the breeders to raise more or higher quality offspring successfully. In the facultatively cooperative‐breeding Tibetan ground tit Pseudopodoces humilis, previous studies found that helpers improve total nestling provisioning rates and fledgling recruitment, but have no apparent effects on the number and body mass of fledglings produced, while breeders with helpers show reduced provisioning rates and higher survival. Here, we investigated whether some of these effects may be explained by female breeders reducing their investment in eggs in response to helper presence. In addition, we investigated whether egg investment is associated with the female breeder's future fitness. Our results showed that helper presence had no effect on the female breeders’ egg investment, and that egg investment was not associated with breeder survival and reproductive success. Our findings suggest that the responses of breeders to helping should be investigated throughout the breeding cycle, because the conclusions regarding the breeders’ adjustment of reproductive investment in response to being helped may depend on which stage of the breeding cycle is considered.     

Behavioural ecology of the North Sulawesi Tarictic Hornbill Penelopides exarhatus exarhatus during the breeding season

Two nests of the North Sulawesi Tarictic Hornbill Penelopides exarhatus exarhatus were monitored for one breeding season (April-July). The females sealed themselves into the nest cavities and remained there for 70–90 days. Incubation was estimated at 16–19 days, and at both nests two offspring fledged. Tarictic Hornbills are cooperative breeders with up to three helpers at the nest and defend foraging territories of 72–139 ha. Food items delivered to the nest included fruits of 34 species (85% of diet items) and invertebrates. Provisioning by helpers allowed breeding males to reduce investment in parental care and may accelerate the development rates in chicks. Constraints on dispersal probably result from habitat saturation rather than living in an unpredictable environment.     

Temporal patterns of circulating LH and steroid hormones in a brood parasite, the Brown-headed cowbird, Molothrus afer.

The breeding behaviour of female Brown-headed cowbirds, Molothrus ater , is remarkably distinct from that of females of most other passerine species. Cowbirds are brood parasites that do not build nests, incubate eggs or feed young. Instead, they have an extended egg-laying stage, and deposit eggs in the nests of hosts within large breeding territories that are defended from other female cowbirds. In this communication, we report the changes in plasma levels of circulating hormones and in morphological characteristics that occur in free-living female Brown-headed cowbirds over a breeding season.
Plasma LH in female cowbirds increases at the onset of breeding and remains elevated for approximately eight weeks, at which time levels decline to basal values. This pattern is similar to that of follicle development, and is consistent with the prolific reproductive output by individuals of this species. Circulating E, levels are quite variable and do not demonstrate any discernible pattern, while plasma B is elevated during the egg-laying period and declines subsequently.
Testosterone titre (T) is highest early in the breeding season, when territorial interactions are generally most frequent. T levels decline slowly thereafter, reaching their lowest values at the season's end. Plasma DHT levels are slightly higher than those for T; DHT exhibits little change until late in the breeding season, when it declines. Fat scores are highest prior to egg laying, and fat reserves diminish as the breeding season progresses. In contrast, body weight increases during the reproductive season, probably as a result of developing eggs in the ovary. These results are consistent with the hypothesis that temporal patterns of hormone secretion are correlated with the expression of social systems.     

Age-specific variation in reproduction is largely explained by the timing of territory establishment in the New Zealand stitchbird Notiomystis cincta

1. Using data from 327 nests over a consecutive 8-year period we examined age-specific variation in reproduction in a population of stitchbirds (or hihi) Notiomystis cincta and related how differences in reproductive performance were linked to the timing of territory establishment and breeding. 2. Across the population all reproductive parameters showed a quadratic relationship with an increase mainly between the first and second breeding season and a decline after the fourth year. A longitudinal analysis showed evidence of senescence by the sixth year in the numbers of chicks fledged and recruited. 3. Reproductive increases between years 1 and 2 were the result of poor-quality females dying after their first breeding season (differential selection hypothesis) in combination with surviving females showing improvements in reproduction in their second year (individual improvement/constraint hypothesis). 4. There was no effect of mate experience or territory quality on improvements in breeding between years. 5. The key variable influencing reproductive output was the timing of breeding. Birds that started laying earlier were more likely to lay multiple clutches in any given season. This was the main difference between first-year and older birds; generally first-year birds initiated egg laying later and consequently laid fewer clutches. 6. Approximately half of all first-year birds did not establish their territory until after the breeding season had begun. This delay in territory establishment resulted in these birds delaying breeding, which resulted in them having a lower reproductive output relative to all other birds. First-year birds that managed to establish their territory before breeding commenced, had similar rates of reproduction as older birds. 7. There was a positive relationship between the timing of territory establishment during a female's first year and her hatching date in the previous breeding season. We hypothesize that this was because late-hatched females were less able to effectively compete for territories against earlier-hatched members of their cohort, and this delayed their establishment and breeding in their first year. Thus, this social constraint is likely to be a major factor driving age-specific reproductive variation in this population.     

OBSERVATIONS OF GROUP BEHAVIOUR AND BREEDING BIOLOGY OF THE YELLOW-BILLED SHRIKE CORVINELLA CORVINA

Yellow-billed Shrikes were found to live in groups throughout the year. Within the group, each member helped to defend the group's territory, warn against predators and feed the breeding female, nestlings and fledglings.
During the study there was little change in the location of the boundaries and in the areas of the territories occupied by the majority of the groups. The densities of the larger groups were in general two to three times that of smaller groups. Numbers within one group varied by ±24% of the average (12) during a period of three years.
Progeny remained in a group for some years before dispersing, sometimes in parties of the same sex. Both sexes exchanged groups, the females moving on average further than males. During successive periods in the history of a group the representation of the sexes varied from a surplus of females to a surplus of males. In the population as a whole the sex ratio was probably parity.
Only one female bred in a group at a particular time and she alone incubated. Eggs were laid on consecutive days. Breeding started at the height of the dry season; the first peak in egg laying occurred at the beginning of the rains; laying continued through the wet season and ceased usually in August. The most frequent clutch size was four, and varied little within a breeding season or between seasons. The incubation period ranged from 15 to 18 days, the most frequently recorded being 17 days. The nestling period was 19 days. The percentage of total eggs laid that produced fledglings was 25% and yearlings 11%.
Young shrikes were independent in the seventh week, participated in group displays in their tenth week and fed fledglings in their fourteenth week.
The age of first breeding was not discovered. Two females in their sixth year were still helpers in a group at the end of the study.     

THE BIOLOGY OF THE WAVED ALBATROSS DIOMEDEA IRRORATA OF HOOD ISLAND, GALAPAGOS

As a nesting species, the Waved Albatross Diomedea irrorata is restricted to Hood Island in the Galapagos archipelago where 12,000 pairs bred in 1971. Outside the islands the species occurs over the northern parts of the Humboldt Current. Two colonies were studied in detail (1970–1971). At the start of a season, males returned first to the colonies and defended a small territory. Copulation occurred without any elaborate ceremony and the female spent little time on land before laying. There was no fixed nest-site, even within a season, and birds moved their eggs considerable distances. This resulted in heavy egg losses. Younger birds bred later than older birds and laid longer but narrower eggs. The average incubation spell varied from four to five days at the extremes of the incubation period to 19 days in the middle. The average incubation and fledging periods were 60 and 167 days respectively. Pairs which lost an egg sometimes adopted the abandoned egg of another bird and successfully reared the chick. Most pairs nested in both seasons. Nesting success was extremely variable, both between years and between colonies. Between 1961 and 1971 at Punta Suarez, virtually no young were reared in four seasons. Even in 1970–71, where nesting success was good, some groups of birds deserted their eggs en masse whereas in neighbouring areas up to 80% of the pairs reared young. The main foods of the young were squid and fish. Birds did not moult wing and tail feathers at the breeding colonies, and about 50% retained some primaries for more than one season, suggesting that successful pairs had difficulty in fitting in a complete moult between breeding attempts. Old feathers were normally found among the inner primaries and at the next moult were preferentially replaced, though adjacent newer feathers were sometimes retained for another season. Some birds bred in their fourth years, but most not until a year or two older. Immatures were present at the colonies late in the breeding cycle, the youngest returning latest and remaining until the last young fledged. Survival of adults and young averaged at least 95% and 93% per annum over many years. Adults and young ringed in 1961 survived equally well. The significance of the timing of the return of immatures and of the large-scale desertion of eggs, apparently not due to food shortage or disturbance, is discussed.