HYBRIDIZATION OF GLAUCOUS GULLS LARUS HYPERBOREUS AND HERRING GULLS L. ARGENTATUS IN ICELAND

Until recently, the Herring Gull Larus argentatus and the Glaucous Gull L. hyperboreus were completely allopatric in western Europe, although having overlapping distributions in Canada and the Soviet Union. But beginning about 1925–1930 a large-scale invasion of argentatus from the Faroe Islands and/or Britain into Iceland occurred, resulting in extensive hybridization with hyperboreus, already present there. The hybrids are viable and apparently fully fertile. In the summers of 1964–1966 371 gulls were collected from breeding colonies in various parts of Iceland and the samples analysed by the use of hybrid indices based on the colour of the five outermost primaries. The populations of northwestern Iceland, where the greatest concentration of hyperboreus was found at the time of the argentatus invasion, were found still to consist largely of pure hyperboreus, with 18–23% of the birds showing signs of argentatus admixture. In eastern Iceland, where argentatus first appeared in numbers, and along the south coast, birds of hybrid origin predominate, 3–43% of the birds being pure argentatus while pure hyperboreus are absent except for one colony in southeastern Iceland where they made up 23% of the birds. Significant differences in composition are found among several of the colonies of eastern and southern Iceland, and possible reasons for this are discussed. Among 53 mated pairs no deviation from random matings with respect to hybrid indices was indicated. The high proportion of pure argentatus found in some colonies in spite of this indicates that pure argentatus are still immigrating from abroad. Two colonies in eastern Iceland sampled both in 1965 and 1966 showed a significant change in composition towards argentatus, further evidence for continued immigration of argentatus. A few other colonies have changed in the same direction over longer periods. Although there must be some differences between the two forms in ecological requirements as shown by their distributions, their ecologies appear similar in Iceland. Both forms take a large portion of their food from the intertidal zone of the sea-shore, and although a slight difference in the selection of foods from the shore was seen in winter in one locality, this probably did not have a genetic basis. However, argentatus may feed to a larger extent on refuse than hyperboreus. This may enable argentatus-like birds to occupy the relatively barren (with respect to intertidal foods) east coast of Iceland in numbers. An examination of several hundred museum skins of argentatus and hyperboreus from other areas shows that occasional hybridization occurred in western Europe long before the extensive contact was established in Iceland, but gene flow between the two was always limited as their ranges were widely separated. The two gulls are effectively reproductively isolated from each other where synipatric in Canada, while the situations in Alaska and in parts of the Soviet Union are unclear. Since the area of overlap without hybridization is much larger than the area of hybridization, the two forms are best regarded as separate species. Future studies may show in which direction the hybrid situation in Iceland will develop, but meanwhile it is suggested that the hybrid populations in which birds of hybrid origin amount to 50% or more be referred to as Larus argentatus/hyperboreus.     

THE FOOD OF SOME LARUS GULLS

The food of L. argentatus on Skomer and Skokholm Islands (Wales) was studied by the analysis of stomach contents, pellets and direct observation. L. argentatus is mainly a scavenger and exists for the most part on fish-waste and garbage discarded by man. At Ncwborough Warren argentatus feeds more on arable land and on the shore than do the Skokholm and Skomer birds.
Although argentatus frequently eats limpets Patella spp., there is little competition for food between them and Oystercatchers Haematopus ostralegus because the two species eat different sized limpets.
A summary of previous studies into the food of argentatus is given.
L. marinus is mainly a predator feeding on the young of other gulls, sea-birds and Rabbits but it also eats carrion, offal and fish-waste when available and during the winter. The numbers of Manx Shearwaters Procellaria puffinus killed on Skokholm and Skomer are discussed.
L. fuscus obtains most of its food on the shore and on arable land. There appears to be little competition, at least in the areas studied, between .fuscus and argentatus for food.     

THE MOULT OF REMIGES AND RECTRICES IN GREAT BLACK-BACKED GULLS LARUS MARZNUS AND GLAUCOUS GULLS L. HYPERBOREUS IN ICELAND

The moult of primaries, secondaries, and rectrices in two closely-related gulls, the Great Black-backed Gull Larus mavinus and the Glaucous Gull L. hyperboreus, was studied in Iceland. Both gulls moult their primaries in an extremely regular sequence, starting with the 1st (innermost) and ending with the 10th (oiltermost) feather. Usually two, less often one or three, primaries are growing per wing during the primary moult, which lasts for about six or seven months. Growlng primaries were estimated to lengthen on the average by 8.7 mm per day in marinus and 7.8 mm per day in hyperboreus. The secondaries, usually 24 in number, are shed in two moult waves, one starting with the innermost feather soon after the start of the primary moult and then progressing slowly outwards, the other beginning with the outermost secondary after the primary moult is about half completed and then progressing rapidly inwards. The moult is completed just before the end of the primary moult as the two moult waves meet at about the 16th secondary. There are no marked differences between the two gulls in the moult of secondaries. The moult of rectrices shows large variations in both species, some feathers being much more irregular than others in their time of shedding. In both species, indications of an obscured centrifugal pattern of replacement are seen, although the 5th (next to the outermost) rectrix is usually the last one to be shed. Significant differences were observed between the two species in the degree of regularity of shedding of some feathers and in the average position in the moulting sequence of others. The moult of rectrices starts soon after the moult of primaries is half completed. The feathers are then shed in rapid succession, and the moult is completed some time before the end of the primary moult. The need for good powers of flight at all times is undoubtedly the reason for the protracted primary moult. This in turn causes the moult to start early, in adults sotnetimes before the eggs are laid; immatures moult even earlier than this. The rectrix moult and the main part of the secondary moult do not begin in adults until the young have fledged, but then progress very rapidly. Presumably, the loss of some of these feathers would impair the flying ability to an extent sufficient to make it difficult for the gulls to care for their young, while the rapid moult is necessary in order for the replacement of these feathers to be completed by the time the primary moult is over.     

DISEASE IN A COLONY OF BLACK-HEADED GULLS LARUS RIDIBUNDUS

An account is given of the disease factors encountered in a colony of Black-headed Gulls in the breeding season 1956.     

ADAPTATIONS TO CLIFF-NESTING IN SOME ARCTIC GULLS (LARUS)

Various aspects of the breeding of three gulls, Larus thayeri, glaucoides and hyperboreus, which nested in colonies both on cliff ledges and on level ground in the eastern Canadian Arctic, were compared with those of the ground-nesting L. argentatus and the cliff-nesting Rissa tridactyla. As the result of adaptation to cliff ledge nesting, many aspects of the breeding biology of R. tridactyla were strikingly different from those of the ground-nesting European L. argentatus, but the behaviour of L. thayeri, glaucoides and hyperboreus clearly spanned these differences. Cliff-nesting individuals of thayeri and glaucoides were most like Rissa; ground-nesting individuals of these species were most like argentatus. L. thayeri was more like Rissa than was glaucoides. With but few exceptions, both cliff- and ground-nesting individuals of hyperboreus were most like argentatus. The factors responsible for the intra-specific differences between cliff- and ground-nesters of thayeri and glaucoides are not clear. Limited gene exchange between cliff and ground colonies occurs. Because of physical features of the nest, first-laid eggs were more liable to fall from ledges than second or third eggs. L. thayeri and glaucoides have evolved separate mechanisms to cope with this problem. Egg shape was multimodal in thayeri and glaucoides. Long pyriform eggs were less liable to fall from ledges than eggs of other shapes. L. thayeri laid more long pyriform eggs as first eggs than did glaucoides. L. thayeri lost fewer eggs than did glaucoides, but glaucoides replaced all lost eggs while thayeri did not. Delayed follicular atresia provided glaucoides with insurance of egg replacement. In thayeri, accessory follicles were reabsorbed after the first egg was laid; in argentatus, after the second egg, and in glaucoides after the third egg. At the approach of a predator, it was advantageous for cliff-dwelling chicks to remain motionless but for ground-dwelling chicks to flee their nests and to hide. Among the cliff-nesters, the “freezing in place” reaction of chicks was best developed in thayeri, to a lesser extent in glaucoides, and least in hyperboreus. Among the ground-nesters, chicks of glaucoides and hyperboreus behaved like those of argentatus and fled their nests when disturbed, but chicks of thayeri froze like their cliff-dwelling siblings. Reciprocal transfers of eggs and chicks between cliff and ground colonies indicated that in argentatus, glaucoides and hyperboreus, the factors determining a chick's reaction to disturbance came into play between hatching and the eighth day. In thayeri, the reaction appeared to be effectively innate. Chicks of glaucoides showed a greater predisposition to this behaviour than chicks of argentatus after both had received identical experience on cliff ledges. In thayeri, stereotypy of the freezing reaction has probably been a factor limiting the colonisation of areas where cliffs are scarce but predators present. In argentatus, lack of perfection of this behaviour (compared to glaucoides and thayeri) has probably been a factor preventing argentatus from attaining cliff ledges. L. hyperboreus, although nowhere abundant, is a widespread species nesting on level ground and on cliff ledges but lacking the modifications observed in glaucoides and thayeri; this is due to its size and aggressiveness, the fact that it picks nest-sites before glaucoides and thayeri arrive in the colonies, and that on cliffs it chooses the largest and most level ledges.     

COLONY DESERTION AND REPRODUCTIVE SYNCHRONY OF BLACK-BILLED GULLS LARUS BULLERI

Black-billed Gulls breeding in large dense colonies on river beds in New Zealand were found to have a high rate of colony desertion early in the nesting cycle. The hypothesis is developed that individuals faced with possible desertion after they have begun to nest can reduce the likelihood of being deserted by delaying their nesting activities until a large number of others is also ready to nest. This interpretation is developed into a graphic model to show that colony desertion can thereby produce a strong selection pressure favouring highly synchronized nesting. This model successfully predicts onset synchrony, subcolony formation, occurrence of complete colony desertions mainly early in the season, occurrence of partial colony desertions, and the existence of nester non-nester conflict over whether or not to desert. Unstable food appears to be the most likely cause of colony desertion in Black-billed Gulls. Black-billed Gulls are similar, in varying degrees to Sandwich, Royal and White-fronted Terns. Convergent traits include tendencies to occupy colonies rapidly and synchronously, reduced site tenacity, and the use of open and often unstable nesting habitats. The links described here between desertion and synchrony may be relevant for some other species.     

SOCIAL STIMULATION OF REPRODUCTION IN HERRING AND LESSER BLACK-BACKED GULLS

The date of the first egg laid and the position of nests were recorded for all pairs in 10 study areas within the mixed Herring and Lesser Black-backed Gull colony on the southern end of Walney Island, Lancashire. The study areas differed from one another in density of breeding birds and in proportion of each species. These data were collected in order to determine if evidence could be found for synchrony of reproduction by social stimulation (the "Fraser Darling Effect"). The gulls in different study areas showed considerable variability in onset and spread of laying; however, attempts to correlate a number of variables, such as density of breeding birds in each area with median date or spread of laying, average nearest neighbour distance with median date of egg-laying, and median laying date for one species with that of the other species in the same section of the gullery, failed to produce convincing evidence for inter- or intraspeciiic social stimulation of reproduction.     

KLEPTOPARASITISM OF SANDWICH TERNS STERNA SANDVICENSIS BY BLACK-HEADED GULLS LARUS RIDIBUNDUS

The parasitic behaviour of Black-headed Gulls in a mixed colony of terns and gulls at the Sands of Forvie on the northeast coast of Scotland is described in some detail. Food-stealing occurred with varying frequency throughout the breeding season. Less than 6% of the terns were attacked during incubation and when their chicks were a few days old, but up to 29% were attacked thereafter. At the same time the percentage of successful attacks rose from 1% or less to 6.5%.
Food items brought back to the ternery were predominantly sandeels, clupeids and gadoids. Only a small proportion of terns carrying fish shorter than 7 cm were attacked whether they were sandeels, clupeids or gadoids.
Robbing success was higher with clupeids and gadoids than with sandeels.
The effect of the Black-headed Gulls' kleptoparasitism on the Sandwich Terns' breeding success is thought to have been negligible during incubation and early chick-life, but might have influenced fledging weight and ultimately post-fledging survival.     

INTRASPECIFIC PREDATION AND COLONIAL BREEDING IN LESSER BLACK-BACKED GULLS LARUS FUSCUS

The breeding biology of Lesser Black-backed Gulls was studied on Skokholm Island, Pembrokeshire, where the number of breeding pairs has been increasing at about 20% per annum since 1963. Laying was found to be synchronous within small groups. Clutch size and breeding success showed seasonal declines over the spread of breeding. The loss mainly of eggs, but to a lesser extent of chicks also, caused this overall decline in success. Hide observations indicated that the bulk of these losses arose through predation by nesting adult Lesser Black-backed Gulls on their nearest neighbours. Not infrequently the protagonist had lost its own clutch shortly before turning predator, and such a chain-sequence should lead to a steady build-up of aggressive failed breeders and so account for the observed seasonal increase in egg loss. Attentiveness to the clutch decreased with season but this was unlikely to have been important in precipitating the predation. Intermediate plant cover was associated with highest nest density and also with highest chick survival. In addition, nest density was directly correlated with chick survival. Whether plant cover and nest density separately affect chick production remains unresolved. Nevertheless, the increasing nest densities in this colony caused by the growth of the gull population are thought to be responsible for the widespread intraspecific predation, the intensity of which is probably a new feature of the gulls' breeding behaviour. The implications of this 'internal' predation for laying synchrony and aggregated nesting are discussed; these two factors of the breeding pattern probably evolved largely to combat 'external' predators. Not only are they no protection against inter-neighbour predation but appear to facilitate it. It remains to be seen whether this kind of predation will significantly affect breeding patterns with further increases in nest density.     

FACTORS AFFECTING THE EGG SIZE OF RED-BILLED GULLS LARUS NOVAEHOLLANDIAE SCOPULINUS

The factors influencing the egg size of the Red-billed Gull Larus novaehollandiae scopulinus were studied at Kaikoura, New Zealand, between 1964 and 1972. In two- and three-egg clutches there was a trend for the eggs to become smaller in the sequence of laying. Length, breadth and volume of eggs of one-, two- and three-egg clutches declined significantly as the season progressed. The size of eggs from single-egg clutches tended to be smaller than eggs from two-egg clutches laid at the same time. There were correlations between the proportions of one-egg and of three-egg clutches being laid at a given period and the mean egg volume of two-egg clutches. When the mean egg volume of two-egg clutches increased there was a corresponding increase in the proportion of two- and three-egg clutches laid. When the mean egg volume of two-egg clutches decreased there was an increase in the proportion of single-egg clutches laid. The egg size of the Red-billed Gull showed no direct correlation with the abundance or availability of food; the largest eggs were produced early in the season when food was in short supply. In spite of an increase in the food supply in the middle of the breeding season, birds laying at this time produced smaller eggs than birds which laid earlier in the season. However, early breeders which relayed at the peak in food abundance on average produced a larger replacement clutch than originals laid early in the season. It is suggested that the birds nesting early in the season are able to produce the largest eggs because they are the most efficient foragers for food, and those which nest later in the season produce smaller eggs, even at peak food abundance, because of their inefficiency or inexperience. Early breeders laying replacement clutches tended to lay larger eggs and larger clutches than birds which are producing their first clutches at the same time. Two-year-old females laid eggs which were significantly shorter than older aged birds while the breadth and volume of the egg increased with the age of the female up to the fifth year. There was a trend for females to lay larger eggs when mated with older rather than younger males. No statistical differences in egg size were detected between females changing or retaining the partner of the previous season. Female body weight and egg volume were positively correlated in females weighing less than 275 g but not for heavier females. It is suggested that the seasonal decline in egg size and clutch size results from a decrease in the availability of food and the ability of the individual to exploit the resource.