Death Of Chickens Research Articles

Survival analysis of Little PenguinEudyptula minorchicks on Motuara Island, New Zealand

Chick survival of Little PenguinsEudyptula minorwas studied on predator‐free Motuara Island, Cook Strait, New Zealand (41d̀05'S, 174d̀15'E), in 1995 and 1996. We used the Kaplan‐Meier estimator and robust Cox regression to estimate chick survival rate (pL se) at 0.325 pL 0.044, leading to an estimated survival from laying to fledging of 0.13 or a reproductive output of 0.26 chicks per pair and breeding attempt. Starvation posed the greatest mortality risk, followed by unknown factors and rain. Risk of death due to rain was restricted to the guard stage, whereas starvation occurred throughout the nesting period, though with a peak in the early guard stage. Significant seasonal differences in survival rate were detected in both years, but with reversed trends, survival decreasing with the season in 1995 and increasing in 1996. Failure of adults to relieve their partner on the nest after chicks hatched accounted for 16% mortality or 34% of all chick deaths. Differences in chick survival rate between nest types were significant in 1995, a year with high rainfall, but not in 1996. Nests in the base of hollow trees had the highest chick survival rate. Of chicks in open nests ‐ a nest type that is unusual for this species ‐ 5.4% fledged. Our results suggest that on Motuara Island good breeding sites are scarce and that the food supply has been poor during the years of this study.

Context and Frequency of Infanticide in Communally Breeding Guira Cuckoos

AbstractWe studied the context of brood reduction through infanticide by communally breeding Guira Cuckoos (Guira guira) in central Brazil. During seven reproductive seasons, we monitored 142 nests from egg laying until fledging. Almost all nests (97%) lost eggs through ejection, and chick deaths occurred in 72% of all nests with hatchlings. There was evidence for infanticide in 38% of the nests that exhibited some mortality. We compared egg and chick mortality in the early part of the season with the later part, when insect abundance declines, but found no significant differences. Less than one-third of all nests monitored showed asynchronous hatching of eggs, and in those that did, chick death was not in reverse hatch order. Although there are several plausible explanations for infanticide, we highlight one likely candidate, which is its interpretation as a sexually selected trait where individuals gain reproductive benefits by provoking the group's nesting failure.

CONTEXT AND FREQUENCY OF INFANTICIDE IN COMMUNALLY BREEDING GUIRA CUCKOOS1

Abstract We studied the context of brood reduction through infanticide by communally breeding Guira Cuckoos (Guira guira) in central Brazil. During seven reproductive seasons, we monitored 142 nests from egg laying until fledging. Almost all nests (97%) lost eggs through ejection, and chick deaths occurred in 72% of all nests with hatchlings. There was evidence for infanticide in 38% of the nests that exhibited some mortality. We compared egg and chick mortality in the early part of the season with the later part, when insect abundance declines, but found no significant differences. Less than one-third of all nests monitored showed asynchronous hatching of eggs, and in those that did, chick death was not in reverse hatch order. Although there are several plausible explanations for infanticide, we highlight one likely candidate, which is its interpretation as a sexually selected trait where individuals gain reproductive benefits by provoking the group's nesting failure.

Determinants of reproductive success in the Hoopoe Upupa epops, a hole-nesting non-passerine bird with asynchronous hatching

Factors affecting success of individual clutches and reproduction in the Hoopoe Upupa epops were studied over five years in a colour-ringed population in Granada, southern Spain. Pairs initiated between one and three breeding attempts per season (including replacement clutches), but only 19% laid a second clutch after raising one successful brood. Seasonal productivity ranged from none to eight fledglings. The pairs that laid a second clutch obtained slightly greater productivity than those that did not, although the differences were not significant, and productivity was not correlated with the number of attempts initiated. Apparently, the reason that these differences were not greater, was the higher predation rate suffered by second and replacement clutches. Predation accounted for 55% of nest losses, although two other major causes were nest desertion and death of females in the nest (17% each). The modal laying date of first clutches in the population was close to the date when the productivity of first clutches was higher. This suggests that, for most individuals, the costs of early laying exceed the benefits of obtaining two broods due to differences in quality between them. Successful clutches produced one to six fledglings (mean 2.97) and the proportion of eggs per clutch that failed to produce fledglings in these successful clutches was very high (49%). Most losses in successful clutches were due to death of chicks, which normally died very young and in a sequence determined by the hierarchy in the brood due to complete hatching asynchrony. The number of chicks fledged was positively correlated with the amount of food carried to the nest by parents. The patterns of nestling mortality and food delivery suggest that the Hoopoe is a brood reduction strategist that reduces ‘optimistic’ clutch sizes by selective starvation of the youngest chicks through extreme hatching asynchrony.

Sibling aggression, hatching asynchrony, and nestling mortality in the black kite ( Milvus migrans )

In siblicidal species, hatching asynchrony could act to reduce sibling rivalry or promote the death of last-hatched chicks. The pattern of hatching asynchrony was experimentally altered in the black kite Milvus migrans. Hatching asynchrony in control broods was intermediate between those of experimentally synchronised and asynchronised broods. Sibling aggression and wounds on the chicks were more commonly observed early in the nestling period and in synchronous nests. Serious injuries were observed on last-hatched chicks in asynchronous nests, as were observations of intimidated or crushed chicks. Sibling aggression was related to food abundance, but some chicks died at an early age in nests with abundant food (cainism). Cainism was more commonly found in asynchronous nests. For species with facultative siblicide, moderate hatching asynchrony could be a compromise between reducing sibling rivalry and avoiding large size differences between sibs that would result in cainism. Female black kites preferentially fed the smallest chicks and exhibited behaviours to reduce sibling aggression, contrary to observations in other siblicidal species. In a highly opportunistic forager such as the black kite, a strategy may exist to protract the life of all the chicks in the brood, waiting for unpredictable situations of food overabundance. This would induce the appearance of a parent-offspring conflict over brood reduction, reflected in the existence of a possible anticipated response by some of the chicks (cainism) and in the appearance of special behaviours by the parents to selectively feed smaller chicks or reduce sibling aggression. In this facultatively siblicidal species, cainism does not seem to be the final stage of an evolutionary trend favouring the raising of high-quality chicks, but a manifestation of a parent-offspring conflict over brood size.

Effects of habitat and introduced mammalian predators on the breeding success of Yellow-eyed Penguins Megadyptes antipodes, South Island, New Zealand

The endemic Yellow-eyed Penguin Megadyptes antipodes is threatened by habitat loss and introduced predators on mainland New Zealand. Nine colonies in the Catlins (south-east coast of South Island) were studied to measure breeding success, penguin abundance, and predator abundance in three successive breeding seasons (1991/92 to 1993/94). Nest numbers increased in all nine colonies in the three years despite predation (probably by Stoats Mustefa erminea) being the most important cause of breeding failure. Larger colonies with higher breeding success were in small gullies with limited shrubs and bushes rather than in the most intact mature forest colonies hitherto assumed to be optimal habitat for the birds. Penguin nests were concentrated near the forest edge, but predators were not, so the predation risk was not elevated near the forest edge. Fragmentation of the original forest habitat had no observable adverse effect on breeding success. Stoats dominated the predator guild, while Ferrets M. furo and Feral Cats Felis catus were rare. Trapping to kill predators early in the season had no marked effect on subsequent predation losses, but trapping intervention when a predation outbreak occurred curtailed further chick deaths. A simple population model predicts that Yellow-eyed Penguins populations will grow provided the average total chicks loss is less than 43% per season, or at least 0.85 chick per nest fledges each year. This requires predation losses to be less than 34%.

Comparative nesting and feeding ecology of skylarks Alauda arvensis on arable farmland in southern England with special reference to set‐aside

1. A study of skylark Alauda arvensis L. breeding ecology in relation to crop type was carried out from April to August 1992 on arable land in southern England. Set‐aside land was included in this comparative study. 2. Territory density averaged 0·15 ha−1. It was 2–3 times higher in fields of set‐aside and grass, especially permanent pasture, than in winter and spring‐sown cereals. 3. Territory size was nearly twice as large in fields of winter cereals (4·5 ha) than in other crop types (2·5 ha). Where set‐aside was present on one farm, territory size in set‐aside (1·7 ha) was a third lower than in cereals and grass. 4. Nesting began in set‐aside and permanent pasture in April and peaked in late May. Nesting was not detected in spring barley until late May and in silage grass until early June. The density of successful nests in set‐aside fields was more than double that in any of the arable crop types. 5. Average clutch size at hatching was 3·91 eggs in fields of set‐aside, over 15% higher than in silage grass (3·40) and in spring barley (3·27). 6. Fledging success did not differ according to crop type, but productivity, expressed as the number of fledglings produced per hectare, was 0·50 in set‐aside, 0·13 in silage grass, and 0·21 in spring barley. Nests with chicks were not found in fields of winter cereals. The causes of chick death were thought to be predation in set‐aside fields, farming practices in silage grass fields, and suspected starvation in spring cereals. 7. The potentially high nesting success of skylarks in set‐aside implies that sympathetic set‐aside management could play an important part in reversing its decline across the European Union.

Status and management of the Palila, an endangered Hawaiian honeycreeper, 1987-1996

A single, relictual population of Palila Loxioides bailleui, a Hawaiian honeycreeper, survives on the slopes of Mauna Kea volcano on the island of Hawai'i, where it feeds principally on flowers and green seeds of the mamane tree Sophora chrysophylla. The Palila was listed as an endangered species by state and federal governments because of continuing damage to its habitat by browsing Feral and Mouflon Sheep Ovis aries and O. musimon and Goats Capra hircus and because of the bird's restricted range and low numbers. Ecology of the Palila was studied from 1987 to 1996. Annual population estimates fluctuated between 1 600 and 5 700 and averaged 3 400 birds. Estimates varied with availability of mamane seeds, which are less abundant in drought years. In drought years, most birds did not attempt to breed, and survival rates were lower because of a shortage of food. Availability of mamane seeds also showed large seasonal variability. While some nests were preyed upon by Owls Asio flammeus, Cats Felis catus and Rats Rattus rattus, losses were high at the end of the season from unexplained death of eggs and chicks. Genetic studies did not implicate inbreeding depression. Neither avian malaria nor avian pox appeared at this site, where the mosquito vector was absent. However, weather and food shortage worsened towards the end of the nesting season. Availability of food and habitat remain the principal factors limiting increase in the Palila population. Recovery efforts now focus on reducing numbers of feral ungulates, fire management, removing mammalian predators, and developing techniques for captive propagation and introduction to currently unoccupied sites within the bird's former range. Reforestation adjacent to the Mauna Kea Forest Reserve would allow the Palila population to expand and grow.

The performance of farmed ostrich chicks in eastern Australia

A prospective observational epidemiological study was undertaken in the south-eastern region of Queensland in eastern Australia to collect accurate information on the performance of farmed ostriches, and to identify the most important constraints facing on-farm production. This paper (the third in a series of three) focuses upon the performance of 394 chicks that hatched on 11 farms in this region from eggs laid between 1 July 1993 and 30 June 1994. Each chick was observed from hatch until dying, leaving the farm of origin or reaching 4 months of age (whichever occurred first). A total of 60.8% of the chicks survived to 4 months of age, with an overall crude mortality rate during the period of observation of 14.0 deaths per 100 chick-months at risk. The most common causes of death were the development of a lower-limb deformity (most frequently tibiotarsal rotation which accounted for 36% of all chick deaths), ‘fading chick syndrome’ (13%), and salmonellosis (11%). Only 83% of the chicks remained free of tibiotarsal rotation during the period of observation. This condition was generally first detected in chicks between 2 and 10 weeks after hatch, and the median survival time following diagnosis was 10 days. ‘Fading chick syndrome’ mainly affected chicks less than 3 weeks of age and did not appear to spread in a contagious manner. Chicks died following infection with Salmonella typhimurium on only one farm. Chick-level factors affecting survival during the 4 months following hatch were examined using the Cox proportional hazards regression model. After accounting for farm-level effects, risk factors for death or euthanasia during the first 4 months following hatch included the weight of the chick at hatch, and the development of tibiotarsal rotation. The need of the industry-wide development and adoption of objective measures of productivity is discussed, and some relevant measures are proposed.

Effects on PCBs on liver ultrastructure and monooxygenase activities in Japanese quail.

The effect of environmental pollutants such as PCBs and DDT on avian species is well documented. It is proven that chronic high level PCB intoxication perturbs calcium metabolism in birds, affecting eggshell thickness. PCBs have an impact on the liver. which accumulates high levels of toxicants. These induce drug-metabolizing enzyme activities in quail (Coturnix coturnix), herring gull (larus argentatus), and partridge (Prdix perdix). As these enzymes can degrade endogeneous molecules such as steroids, xenobiotics like PCBs can severely hinder birds` reproductive performance. PCBs induce damage such as regression of the testes, decreased sperm concentration, and altered embryonic development resulting in death or malformation of chicks. More ever, ultrastructural alterations linked with induction of these enzymes have been observed in the livers of PCB-contaminated chickens and ducks. This study examines the effects of Aroclor 1254 on liver morphology and glycogen content in quail, and related morphological modification to liver monoxygenase activities. 26 refs., 6 figs., 1 tab.

Rous Sarcoma Virus (Subgroup A) Induced Tumourigenesis in Visceral Organs of Avian Species

Abstract Rout, P. K., Pani, P. K. and Naithani, S. 1993. Rous sarcoma virus (subgroup A) induced tumourigenesis in visceral organs of avian species. J. Appl. Anim. Res., 3: 11–18. An investigation was conducted using four avian species (indigenous chicken, Bantam, quail and guinea fowl) to ascertain Rous sarcoma virus (Subgroup A)-induced tumour expression in animals that had been inoculated as embryos into the chorioallantoic membrane (CAM). It was observed that as little as one virus particle could induce liver tumour (LT) death of chicks. The guinea fowl invariably and equally developed tumours in liver and spleen but did not express tumours in heart in contrast to chickens. The degree of CAM infection was not linearly related to LT deaths in all avian species as even chicks from the group with a zero pock count, had expressed LT incidence. The quail and Bantam embryos were observed to be more sensitive than the chick dying early after inoculation with the same dose of virus.

Effects of Hatching Order, Sibling Asymmetries, and Nest Site on Survival Analysis of Jackass Penguin Chicks

We used survival analysis to examine the fates of Jackass Penguin (Spheniscus demersus) chicks on Dassen Island, off the west coast of South Africa. There were two distinct phases of chick loss. The first, when chicks were 0-34 days old, involved losses primarily due to burrow collapse, exposure and drowning, and accidental death in the nest. This mortality was influenced strongly by nest-site characteristics. Overall reproductive success was lowest in open nests, intermediate in burrows, and highest in rock nests. The probability of chick death due to flooding was highest for burrows in shell/guano conglomerate, whereas the risk of burrow collapse was highest in sand. Burrows in high-density colonies had a greater likelihood of collapsing after heavy rain than burrows in low-density areas. The second phase occurred 42-90 days after hatch, when losses were almost entirely due to starvation. Comparisons of survival in control and experimental nests with varying degrees of sibling asymmetry indicate that it is the extent of hatching asynchrony that affects the ability of the smaller sibling to compete for food and that will produce sibling differences in the risk of starvation. Chicks in two-chick control nests starved at higher frequencies than chicks in less asynchronous experimental broods.

Intraspecific Predation and Chick Survival: Comparison among Colonies of Slaty-Backed Gulls

Relationships between the occurrence of intraspecific predation, population size and nesting density were studied in slaty-backed gulls Larus schistisagus on Teuri, Daikoku and Yururi Islands around Hokkaido, Japan. Chick survival on Daikoku was higher than that on Teuri and Yururi. The main causes of death of chicks were intraspecific predation on Teuri and attacks by neighbor adults on Yururi. Slatybacked gulls frequently preyed upon conspecific chicks on Teuri, where the population size and nesting density were the smallest, while they rarely did on Daikoku, where the population size was the largest and the nesting density was medium and on Yururi, where the nesting density was the highest and the population size was medium. Therefore, intraspecific predation was not a density dependent mortality factor of chicks, hence, could not regulate population dynamics. The regional difference in the occurrence of intraspecific predation did not always relate to breeding synchrony or to chick guarding by parents. Twenty percent of individual male gulls probably were responsible for most of the intraspecific predation. Reproductive success of pairs including these cannibal males was not significantly different from that of non cannibal pairs including males that fed their offspring with seabird chicks but did not prey upon conspecific chicks. Therefore, pairs including cannibalistic males possibly did not have selective advantage over others.

Research Note: Application of the Filial Imprinting Phenomenon to Broiler Chicks at a Commercial Farm

Ten thousand broiler chicks on commercial farms in Israel were imprinted to blue and red plastic boxes (20 × 20 × 40 cm) containing loud speakers (8 Ω, 12 W) equipped with a tape recorder, in which a classical music cassette had been placed. Upon reaching 3 wk of age, an attempt was made to move these chicks from the training area (A) to a new area (B) using the imprinting objects. This experiment was repeated four times: May, August and November of 1985, and in January 1986. During the warm months of May and August 30 to 60% of the experimental chicks followed the imprinting objects toward Area B. During the cold months and in all four repetitions with the controls, less than 5% of the flocks moved from Area A to Area B. Winter chicks may have been imprinted on brooder heaters. In all repetitions, most of the imprinted chicks distributed themselves equally over Area B 1 h after the doors between Areas B and A were closed. Controls, however, crowded around the door between Areas A and B resulting in the death of many chicks by asphyxiation. This study has shown that thousands of heavy strain chicks can be easily attached to an imprinting stimulus and it may be possible to apply this phenomenon for commercial purposes.

Siblicide, parent-offspring conflict, and unequal parental investment by egrets and herons

O'Connor (1978) generated behavioral predictions from Trivers's (1974) theory of parent-offspring conflict (POC) in relation to avian brood reduction, especially for species using escalated sibling aggression. In general, when daily mortality rates fall between certain values (the “mortality difference thresholds” for parents and surviving offspring: O'Connor 1978), POC is expected. At such times parents and the early-hatching “senior” offspring are expected to disagree on the necessity, or at least timing, of the “victim” chick's death. Because that death is promoted directly by the senior chicks and because parents have realistic means of nullifying those efforts, behavioral conflict is expected. I examined the activities of all family members in one siblicidal (great egrets, Casmerodius albus) and one nonsiblicidal brood-reducing species (great blue herons, Ardea herodias) of ardeidae in Texas nesting colonies. Daily mortality rates over two breeding seasons approximated the predicted POC zone for four-chick (“B/4”) heron broods and three-chick (B/3) egret broods. As predicted, egret sibling aggression was significantly lower in two-chick broods. Herons seldom fought, regardless of brood size. Special attention was paid to the egret B/3 data for behavioral manifestations of POC. Parents did not interfere overtly with sibling fights (99.2% of 2,829), but parents seemed to have a subtle mollifying effect on the aggression (fewer “sever” fights occurred when parents were present). Parents showed no favoritism in food distribution to dying victim chicks. The mixed results are consistent with two rather different POC interpretations: (1) that significant conflict exists and the offspring are “winning” or (2) that there is no significant evolutionary conflict. In the latter case, parents are regarded as creating asymmetries among offspring (e.g., via hatch asynchrony), then leaving competition among siblings to run its course without further intervention (“laissezfaire” policy). This view explicitly recognizes that siblicide and other investment-skewing selfish chick behaviors may serve parental interests well-i.e., that equal allocation of resources is not likely to be the parental optimum. Using the relative reproductive values of each nestling as an index of the parent's unequal parental investment optimum produces a closer fit to the observed food distribution skew, but does not resolve the question of whether conflict remains.

Brood reduction in the Blue‐eyed Shag Phalacrocorax atriceps

Brood reduction is common in a population of Blue‐eyed Shags on Signy Island, South Orkney Islands. This paper describes possible adaptations which may reduce the brood. In clutches of three, the last egg was smaller, and hatched 2.4 days later than its siblings. Whilst 78–84% of first and second (‘A’ & ‘B’) chicks fledged, only 11 % of ‘C’ chicks did. In a sample of artificially synchronized broods chick survival was as high as in normal asynchronously hatching broods, but there were more cases of total brood loss. The age at which the C chick died was related inversely to the length of the A‐C hatching interval. Relative differences in sibling weights were highest during the first 12 days, when most of the C chick deaths occurred. At this age the daily food requirements of each brood of three was one‐tenth that of each brood of two just prior to fledging. It is suggested that C chicks were unable to compete effectively for a food supply which was limited by the parents, rather than by the environment. The asymptotic weight attained by A chicks was inversely related to brood size, and was greater than that of B or C chicks. Normal asynchronous broods produced at least one heavy (A) chick and one medium weight (B) chick, whilst in synchronized broods the asymptotic weight attained was similar to that of B chicks in normal broods.

Maternal nutrition, egg quality and breeding success of Scottish Ptarmigan Lagopus mutus

Samples of Ptarmigan eggs were hatched in captivity. They had been taken from ground adjacent to study areas where the wild birds' breeding success by early August, and their food plants, were measured.Almost all deaths of chicks in captivity occurred within a few days of hatching and examination postmortem showed no specific cause. The proportion of chicks which died in captivity before 15 days of age varied markedly between years and study areas. These variations occurred in parallel with variations in breeding success in the wild populations from which the eggs had been taken. It was concluded that the survival of chicks both in captivity and in the wild was determined in part before the eggs hatched.Variations in breeding success from year to year were correlated with the number of days that the food plants had been growing before the hens finished laying. Breeding success in the wild and chick survival in captivity were better for a ‘rich’ area overlying some base‐rich rocks than for a ‘poor’ one overlying granite and with less blaeberry. It was inferred that, as in Red Grouse, maternal nutrition affected breeding success through the quality of the eggs.

The chick as a model for malignant hyperpyrexia

Administration of succinylcholine (SCh) to chicks produces rigid paralysis and death due to respiratory impairment. The mechanism of the SCh effect is probably related to the multiple innervation of muscle fibres, leading to excessive intracellular accumulation of calcium. This situation may be similar to that in malignant hyperpyerxia (MH) occurring in mammals. Dantrolene sodium, phenytoin and procain, drugs used against MH, were found to afford protection against SCh rigidity and death in chicks. It is suggested that the chick can be used as a convenient model for a rapid screening of drugs potentially active against MH.

Ticks, Louping Ill and Red Grouse on Moors in Speyside, Scotland

The number of nymphal ticks (Ixodes ricinus) on red grouse (Lagopus lagopus scoticus) chicks was inversely related to the adult birds' breeding success and population density. Some chicks less than 14 days old were found apparently blinded and dying probably due solely to very high infestations of tick larvae. In addition, tick-borne louping-ill virus, which is known to kill grouse, was probably an important cause of death of grouse chicks in this study. J. WILDL. MANAGE. 42(3):500-505 Population densities of red grouse dif- fer from area to area, depending largely upon differences in the quantity and quality of their main food, heather (Cal- luna vulgaris) (Watson and Moss 1972). However, some parts of Scotland are in- fested with numerous sheep ticks and Moore (1937) suggested that ticks kill many young grouse. This is also the opin- ion of gamekeepers, who complain that grouse are scarce on moors with many ticks. In this study we tested these sug- gestions by comparing the number of sheep ticks found on red grouse chicks with the mortality rate of the chicks and the breeding success and breeding den- sity of adult grouse. Ticks suck blood and body fluids and cause inflammation at their point of at- tachment. They also transmit louping-ill virus (Flavivirus group), which is patho- genic to red grouse (Reid 1975). Hence the ticks' role as vectors of louping ill may be more important than the harm they do by removing blood and causing inflammation. Ticks occur as larvae,

Breeding Success of the Common Puffin (Fratercula arctica L.) on Different Habitats at Great Island, Newfoundland

Factors contributing to breeding success of puffins were studied on Great Island, Newfoundland, in 1968 and 1969. Puffin burrow density was negatively correlated with distance from the cliff edge and positively correlated with angle of slope. These correlations are biologically significant in that close to the cliff edge, where the angle of slope was steep, breeding success was significantly higher than on adjacent level habitat. In spring both habitats were occupied simultaneously, and nest—site tenacity was equally strong in them. During settlement the frequency of fighting was higher and the peak was reached earlier on slope habitat. Males were heavier on slope than on level habitat just after peak egg laying, although wing lengths were similar. Females were similar on the two habitats. Measurements of eggs from both habitat were the same. Egg—laying dates were also similar, but annual variation was greater on level than on slope habitat, as was variation within a single year. Hatching success was higher on slope habitat, mainly because the incidence of egg disappearance was lower during incubation; also the frequency of infertile eggs was greater on level habitat. Fledging success was higher on slope habitat, and higher on both habitats in 1969. Frequency of chick deaths in the nest and disappearance before fledging was higher on level habitat in both years. Fledging success was higher for early—hatched chicks in both habitats. Total breeding success was higher on slope habitat, and higher in both habitats in 1969. Fledging condition of chicks varied according to the habitat and time period in which they were raised. On the average, body weight at fledging was greater and less variable for birds on slope habitat; early—hatched chicks were heaviest in both habitats. Wing length of young at fledging did not differ. Early—hatched chicks on slope habitat fledged quickest (also quicker than early—hatched chicks on level habitat); late—hatched birds on slope fledged slower than late—hatched birds on level habitat. Age at fledging was greater in 1968 than in 1969 on both habitats. Meal size delivered to chicks by parents was the same on the two habitats, but frequency of feeding was greater on slope than level habitat. Breeding pairs on slope habitat were unable to raise two chicks (artificial twins) to fledging; similarly, one adult could not rear a single chick. Adults feeding chicks on level habitat were attacked and robbed more frequently by gulls than birds on slope habitat, probably because escape (take—off) from an attack was quicker on slope than on level ground. Experiments performed to determine causes of the differential egg and chick loss before fledging in the two habitats showed that during incubation the proportion of eggs displaced to the burrow entrance by incubating birds when leaving the burrow in a hurry was the same, but the rate of panic flights was higher on level habitat. Also, chicks when starved spent more time at the burrow entrance than when fed regularly. It is concluded that the difference in breeding success on the two habitats was due to a higher exposure of eggs and chicks to gull predation on level habitat. The primary cause for this differential exposure was that adults on slope habitat were less vulnerable to gull disturbance during incubation and gull robbery when feeding their chicks. Thus breeding failures resulted from the interactions of food shortage and gull interference. This conclusion was tested by comparing breeding performance of birds at Great Island with birds at two islands where gull interference was absent. Egg and chick survival was greater under "gull—free" conditions. In addition, body weights at fledging were higher and less variable at the colonies without gull interference than at Great Island. The ways in which natural selection acts upon puffins at the breeding colony are considered; a model relating puffin nest distribution and habitat features is presented, and predictions which might be tested are outlined.