Year‐round spatial movements and trophic ecology of Trindade Petrels (Pterodroma arminjoniana)

Trindade Petrels (Pterodroma arminjoniana) are vulnerable gadfly petrels that breed on the remote Trindade Island, located ~1100 km off the Brazilian coast. Little is known about their spatial ecology, and their trophic ecology has only been described for the breeding season. We tagged four Trindade Petrels with global location sensing loggers (GLS) from October 2013 to November 2014 and sampled the blood and feathers (innermost primary and the eighth secondary) of 14 individuals to evaluate their year‐round spatial and isotopic ecology. We examined individual distributions, habitat use and suitability, activity, and isotopic values during the breeding, migration, and non‐breeding periods. Trindade Petrels used areas in the southwest Atlantic Ocean (between 10°N and 50°S in latitude) during the breeding season. They migrated through pelagic waters of the tropical Atlantic to the northwest Atlantic, where they spent the non‐breeding season. Trindade Petrels used mostly tropical to subtropical waters in areas of intermediate to high wind speeds and low marine productivity. Individuals spent more time foraging at night than during the day. During the breeding season, birds in northerly areas had higher carbon‐13 values, and birds that used more pelagic areas foraged on prey at a higher trophic level (higher nitrogen‐15 values) than those in more southern and coastal areas. Isotopic values during the breeding, migration, and non‐breeding periods differed, possibly due to differences among individuals in their at‐sea distribution throughout the year. We confirmed the non‐breeding distribution of Trindade Petrels, which was previously known only from vessel sightings and stranded birds. Our results also suggest a strong temporal segregation in the at‐sea distribution and trophic ecology between two groups of individuals, which might indicate the existence of two separate breeding populations.     

Year‐round movements of sympatric Fork‐tailed (Oceanodroma furcata) and Leach's (O. leucorhoa) storm‐petrels

Long‐distance movements are characteristic of most seabirds in the order Procellariiformes. However, little is known about the migration and foraging ranges of many of the smaller species in this order, especially storm‐petrels (Hydrobatidae). We used Global Location Sensors to document the year‐round movements of sympatrically breeding Fork‐tailed Storm‐Petrels (Oceanodroma furcata) and Leach's Storm‐Petrels (O. leucorhoa) from the Gillam Islands located northwest of Vancouver Island, British Columbia, Canada. In 2016, breeding Fork‐tailed (N = 5) and Leach's (N = 2) storm‐petrels traveled maximum distances of ~1550–1600 km from their colony to a region that has a wide shelf with major canyons creating a highly productive foraging area. After the breeding season, Fork‐tailed Storm‐Petrels (N = 2) traveled to similar areas west of the Gillam Islands, a maximum distance of ~3600 km from the breeding colony, and remained in the North Pacific Ocean and north of the Subarctic Boundary for an average of 5.4 mo. Post‐breeding Leach's Storm‐Petrels (N = 2) moved south to the Eastern Tropical Pacific, west of central Mexico, Ecuador, and northern Peru, an estimated maximum distance of ~6700 km from their breeding colony, and remained there for an average of 7.2 mo. Carbon (δ¹³C) and nitrogen (δ¹⁵N) stable isotope analyses of feathers revealed niche separation between Fork‐tailed (N = 21) and Leach's (N = 53) storm‐petrels. The wide range of δ¹⁵N values in the feathers of Leach's Storm‐Petrels (N = 53) suggests that they foraged at a variety of trophic levels during the non‐breeding season. Our results demonstrate that storm‐petrels have large core foraging areas and occupy vast oceanic areas in the Pacific during their annual cycle. However, given the coarse precision of Global Location Sensors, additional study is needed to identify the specific areas used by each species during both breeding and non‐breeding periods.     

Breeding strategies of Antarctic Petrels Thalassoica antarctica and Southern Fulmars Fulmarus glacialoides in the high Antarctic and implications for reproductive success

Breeding strategies of two closely related fulmarine petrels were studied on Ardery Island, on the continental coast of East Antarctica, where short summers are expected to narrow the time-window for reproduction. Both species had a similar breeding period (97 days from laying to fledging) but Antarctic Petrels Thalassoica antarctica bred up to 16 days earlier than Southern Fulmars. During the pre-laying exodus, all Antarctic Petrels deserted the colony, whereas some Southern Fulmars Fulmarus glacialoides remained. Antarctic Petrels exhibited stronger synchronization in breeding, made longer foraging trips and spent less time guarding their chicks than Southern Fulmars. Overall breeding success of both species was similar but failures of Antarctic Petrels were concentrated in the early egg-phase and after hatching, when parents ceased guarding. Southern Fulmars lost eggs and chicks later in the breeding cycle and so wasted more parental investment in failed breeding attempts. Different breeding strategies may be imposed by flight characteristics; Southern Fulmars are less capable of crossing large expanses of pack ice and need to delay breeding until the sea ice retreats and breaks up. However, due to the short summer they risk chick failure when weather conditions deteriorate late in the season.     

A further contribution on the biology of the Antarctic flea,Glaciopsyllus antarcticus (Siphonaptera: Ceratophyllidae)

Summary Experiments showed that adult Antarctic Fleas Gladopsyllus antarcticus preferred dark over light conditions, dry over wet conditions, and fine substrate over coarse substrate. Examination of seabird chicks and collections of nest material indicated that the Southern Fulmar Fulmarus glacialoides, is the major host species of the Antarctic Flea, while Snow Petrels Pagodroma nivea, Cape Petrels Daption capense, Antarctic Petrels Thalassoica antarctica, and Wilson's Storm-Petrels Oceanites oceanicus are minor hosts. This is the first report of Antarctic Fleas occuring on either Antarctic Petrels or Wilson's Storm-Petrels. No fleas were found associated with Southern Giant Petrels Macronectes giganteus, Antarctic Skuas Catharacta maccormicki, or Adelie Penguins Pygoscelis adeliae. No live fleas were found in Southern Fulmar nest material during their period of dispersal over the winter months, supporting the hypothesis that Antarctic Fleas survive the winter period by remaining on their seabird host.     

Segregation in space and time explains the coexistence of two sympatric sub‐Antarctic petrels

Biological communities are shaped by competition between and within species. Competition is often reduced by inter‐ and intraspecific specialization on resources, such as differencet foraging areas or time, allowing similar species to coexist and potentially contributing to reproductive isolation. Here, we examine the simultaneous role of temporal and spatial foraging segregation within and between two sympatric sister species of seabirds, Northern Macronectes halli and Southern Macronectes giganteus Giant Petrels. These species show marked sexual size dimorphism and allochrony (with earlier breeding by Northern Giant Petrels) but this is the first study to test for differences in foraging behaviours and areas across the entire breeding season both between the two species and between the sexes. We tracked males and females of both species in all breeding stages at Bird Island, South Georgia, to test how foraging distribution, behaviour and habitat use vary between and within species in biological time (incubation, brood‐guard or post‐brood stages) and in absolute time (calendar date). Within each breeding stage, both species took trips of comparable duration to similar areas, but due to breeding allochrony they segregated temporally. Northern Giant Petrels had a somewhat smaller foraging range than Southern Giant Petrels, reflecting their greater exploitation of local carrion and probably contributing to their recent higher population growth. Within species, segregation was spatial, with females generally taking longer, more pelagic trips than males, although both sexes of both species showed unexpectedly plastic foraging behaviour. There was little evidence of interspecific differences in habitat use. Thus, in giant petrels, temporal segregation reduces interspecific competition and sexual segregation reduces intraspecific competition. These results demonstrate how both specialization and dynamic changes in foraging strategies at different scales underpin resource division within a community.     

Energetic solutions of Rock Sandpipers to harsh winter conditions rely on prey quality

Rock Sandpipers Calidris ptilocnemis have the most northerly non‐breeding distribution of any shorebird in the Pacific Basin (upper Cook Inlet, Alaska; 61°N, 151°W). In terms of freezing temperatures, persistent winds and pervasive ice, this site is the harshest used by shorebirds during winter. We integrated physiological, metabolic, behavioural and environmental aspects of the non‐breeding ecology of Rock Sandpipers at the northern extent of their range to determine the relative importance of these factors in facilitating their unique non‐breeding ecology. Not surprisingly, estimated daily energetic demands were greatest during January, the coldest period of winter. These estimates were greatest for foraging birds, and exceeded basal metabolic rates by a factor of 6.5, a scope of increase that approaches the maximum sustained rate of energetic output by shorebirds during periods of migration, but far exceeds these periods in duration. We assessed the quality of their primary prey, the bivalve Macoma balthica, to determine the daily foraging duration required by Rock Sandpipers to satisfy such energetic demands. Based on size‐specific estimates of M. balthica quality, Rock Sandpipers require over 13 h/day of foraging time in upper Cook Inlet in January, even when feeding on the highest quality prey. This range approaches the average daily duration of mudflat availability in this region (c. 18 h), a maximum value that annually decreases due to the accumulation of shore‐fast ice. Rock Sandpipers are likely to maximize access to foraging sites by following the exposure of ice‐free mudflats across the upper Cook Inlet region and by selecting smaller, higher quality M. balthica to minimize foraging times. Ultimately, this unusual non‐breeding ecology relies on the high quality of their prey resources. Compared with other sites across their range, M. balthica from upper Cook Inlet have relatively light shells, potentially the result of the region's depauperate invertebrate predator community. Given the delicate balance between environmental and prey conditions that currently make Cook Inlet a viable wintering area for Rock Sandpipers, small variations in these variables may affect the suitability of the site in the future.     

Skuas (Stercorarius spp.) moult body feathers during both the breeding and inter‐breeding periods: implications for stable isotope investigations in seabirds

Seabirds are mostly thought to moult during the inter‐breeding period and the isotopic values of their feathers are often therefore assumed to relate to their assimilated diet during such periods. We observed Brown Skuas Stercorarius antarcticus lonnbergi and South Polar Skuas Stercorarius maccormicki moulting on a breeding site at King George Island, Antarctica. This raises concerns about the reliability of using stable isotopes in feathers to infer feeding localities of birds during the inter‐breeding period. We analysed the δ¹³C and δ¹⁵N values of growing and fully grown body feathers collected from the same individuals. For both species, δ¹³C values of growing feathers indicated feeding areas in the Antarctic zone (breeding grounds), whereas most fully grown feathers (100% for South Polar Skuas and 93.3% for Brown Skuas) could be assigned to northern latitudes (non‐breeding grounds). However, a few fully grown body feathers of Brown Skuas (6.7% of the feathers, belonging to two birds) showed isotopic values that indicated moult in the Antarctic zone. As the growth period of those feathers was unknown, they could not be used with confidence to depict the foraging behaviour of the birds during the non‐breeding period. Although precautions must be taken when inferring dietary information from feathers in seabirds where the moulting pattern is unknown, this study shows that if the development stage of a feather (growing/fully grown) is identified, then dietary information from both breeding and non‐breeding seasons can be obtained on the same individual birds.     

An effective method for trapping scavenging seabirds at sea

ABSTRACT Most studies of seabirds that involve trapping and marking birds are carried out at breeding colonies. This bias toward the breeding period and colony‐based research is partially caused by difficulties in capturing birds at sea. From 2005 to 2007, we used a cast net thrown by hand from a fishing boat to capture albatrosses and petrels at sea in the southwestern Atlantic Ocean. About 500 birds of 13 species were captured, ranging in size from the 30‐g Wilson's Storm‐petrel (Oceanites oceanicus) to the 10‐kg Wandering Albatross (Diomedea exulans). Cast nets are potentially useful for capturing any seabird that can be attracted close to fishing vessels by bait, such as sardines, squid, offal, or shark liver, thrown into the water. Our method was most effective for capturing bold species, such as Cape (Daption capense), Spectacled (Procellaria conspicillata), and White‐chinned (Procellaria aequinoctialis) petrels and Great Shearwaters (Puffinus gravis), but was not effective for capturing shy species, such as Cory's (Calonectris diomedea), Cape Verde (Calonectris edwardsii) and Manx (Puffinus puffinus) shearwaters, species that rarely sit on the water, such as Wilson's Storm‐petrels, Thin‐billed (Pachyptila belcheri) and Antarctic (Pachyptila desolata) prions and Atlantic Petrels (Pterodroma incerta), and species with excellent diving abilities, such as Sooty Shearwaters (Puffinus griseus). For many species of seabirds, cast nets would likely be more efficient for capturing large numbers of birds than other methods, such as hoop nets.     

Sexual dimorphism,moult and body condition of seabirds killed by longline vessels around the Prince Edward Islands, 1996–97

Ryan, P.G. 1999. Sexual dimorphism, moult and body condition of seabirds killed by longline vessels around the Prince Edward Islands, 1996–97. Ostrich 70 (3&4): 187–192.

A total of 393 seabirds from nine species killed by longline fishing in South African waters around the Prince Edward Islands during summer 1996–97 were sexed, measured, and their body fat levels and primary moults scored. Males of Whitechinned Petrels Procellaria aequinoctialis and Greyheaded and (Indian Ocean) Yellownosed Mollymawks Thalassarche chrysostoma and T. [chlororhynchosj bassi averaged 1–9% larger than females. Measures of bill depth were the most dimorphic characters in all three species. I recommend that if only a single measure is taken to estimate sex of adults, bill depth at the base be the standard measure for petrels, and bill depth at the nail be the standard measure for mollymawks. Discriminant functions are described for predicting the sex of unknown individuals; the proportion of misclassified birds ranged from 2–10%. With the exception of giant petrels Macronectes spp., only two young birds were actively moulting their primary flight feathers. Analysis of wear patterns among adult Greyheaded Mollymawk primaries conforms with studies of moult in this species at South Georgia. Fat scores did not differ between sexes. They tended to decrease during the breeding season, but trends were masked by great inter-individual variation. There was no significant relationship between fat scores and the presence of ingested plastics. Plastic debris was found in the stomachs of 90 birds from five species. Incidence in Whitechinned Petrel stomachs (37%) was less than that reported from birds collected off South Africa in the early 1980s (57%), but this is probably due to the predominance of breeding adults. The frequency and size of plastic loads decreased through the breeding season, which is consistent with the hypothesis that inter-generational transfer is important in the dynamics of ingested debris.     

Migratory movements and wintering areas of Leach's Storm‐Petrels tracked using geolocators

Accumulating evidence suggests that Atlantic populations of Leach's Storm‐Petrels (Oceanodroma leucorhoa) are experiencing significant declines. To better understand possible causes of these declines, we used geolocators to document movements of these small (～50‐g) pelagic seabirds during migration and the non‐breeding period. During 2012 and 2013, movement tracks were obtained from two birds that traveled in a clock‐wise direction from two breeding colonies in eastern Canada (Bon Portage Island, Nova Scotia, and Gull Island, Newfoundland) to winter in tropical waters. The bird from Bon Portage Island started its migration towards Cape Verde in October, arrived at its wintering area off the coast of eastern Brazil in January, and started migration back to Nova Scotia in April. The bird from Gull Island staged off Newfoundland in November and then again off Cape Verde in January before its geolocator stopped working. Movements of Leach's Storm‐Petrels in our study and those of several other procellariiforms during the non‐breeding period are likely facilitated by the prevailing easterly trade winds and the Antilles and Gulf Stream currents. Although staging and wintering areas used by Leach's Storm‐Petrels in our study were characterized by low productivity, the West Africa and northeastern Brazilian waters are actively used by fisheries and discards can attract Leach's Storm‐Petrels. Our results provide an initial step towards understanding movements of Leach's Storm‐Petrels during the non‐breeding period, but further tracking is required to confirm generality of their migratory routes, staging areas, and wintering ranges.     

UNCOMMON VISITORS TO MARANDELLAS DISTRICT

Schmitt. M. B. 1976: Observations on the Cape Rail in the Southern Transvaal. Ostrich 47: 16–26.

For five years a population of Cape Rails Rallus caerulescens has been studied on a vlei in the southern Transvaal. Morphological differences between male and female are discussed, the breeding season is indicated and the development of chicks is described. In autumn a migration possibly of birds from drier areas invades the study area. The breeding population appears to be resident. The Cape Rail moults all flight feathers simultaneously during its breeding season.     

New insights from an attempt to reintroduce Red‐cockaded Woodpeckers in northern Florida

Red‐cockaded Woodpeckers (Leuconotopicus borealis) were extirpated from Tall Timbers Research Station in the early 1980s. To help meet conservation goals established for this imperiled species in north Florida, we attempted to reintroduce the woodpecker to the research station by constructing artificial cavities and translocating 27 subadult woodpeckers from 2006 to 2010. Successful nesting occurred during the breeding season following the initial translocation of four male–female pairs. Translocations were suspended in 2011 when breeding groups occupied 6 of 12 available clusters of cavity trees. The population increased steadily after 2011 and, in 2015, totaled 28 adults distributed among nine breeding groups plus a single territorial male. The 2015 population included 22 individuals produced at Tall Timbers, an immigrant female, and five birds originally translocated as subadults. Seven breeding groups in 2015 also had non‐breeding helpers. New milestones documented during this reintroduction attempt included recruitment of locally produced birds into the breeding population, excavation of natural cavities, two immigration events, and natural expansion into an unoccupied area. We also documented the threat that heavy rains may pose to small populations. Expenses totaled $211,000 during the first 5 yr when translocations and cavity construction were the primary activities. After translocations were suspended, recurring management expenses were ~ $6500 annually. Because our founding population was small (N = 12), intermittent translocations will likely be needed in the future to offset the deleterious effects of inbreeding.     

Speciation and phylogeography of giant petrels Macronectes

We examine global phylogeography of the two forms of giant petrel Macronectes spp. Although previously considered to be a single taxon, and despite debate over the status of some populations and the existence of minimal genetic data (one mitochondrial cytochrome b sequence per form), the current consensus based on morphology is that there are two species, Northern Giant Petrel M. halli and Southern Giant Petrel M. giganteus. This study examined genetic variation at cytochrome b as well as six microsatellite loci in giant petrels from 22 islands, representing most island groups at which the two species breed. Both markers support separate species status, although sequence divergence in cytochrome b was only 0.42% (corrected). Divergence was estimated to have occurred approximately 0.2 mya, but with some colonies apparently separated for longer (up to 0.5 my). Three clades were found within giant petrels, which separated approximately 0.7 mya, with the Southern Giant Petrel paraphyletic to a monophyletic Northern Giant Petrel. There was evidence of past fragmentation during the Pleistocene, with subsequent secondary contact within Southern Giant Petrels. The analysis also suggested a period of past population expansion that corresponded roughly to the timing of speciation and the separation of an ancestral giant petrel population from the fulmar Fulmarus clade.     

Experimental evidence that a large raptor can detect and replace heavily damaged flight feathers long before their scheduled moult dates

Because many species of large birds must remain capable of flight during moult and breeding, complete replacement of all flight feathers often takes two or more years, with the result that their plumage normally includes many faded, worn and sometimes even broken feathers. It seems adaptive for such birds to have the ability to quickly replace severely damaged feathers. In search of such a feather replacement mechanism, we cut rectrices on a captive Golden Eagle Aquila chrysaetos and found that feathers cut in their first year of use were replaced, on average, after 11.4 months, whereas uncut feathers before and during the experiment were moulted, on average, after 24 months of use. Feather cutting advanced moult, on average, in excess of a year and thereby demonstrates the existence of a previously undescribed neurophysiological mechanism for preferentially replacing damaged feathers.     

State uncertainty models and mark‐resight models for understanding non‐breeding site use by Piping Plovers

Conservation of beach‐nesting medium‐distance migrants has focused on breeding areas because protection of nests is more tractable than protection of non‐breeding habitat. As breeding ground management has encountered diminishing returns, interest in understanding threats in non‐breeding areas has increased. However, robust estimates of non‐breeding demographic rates and abundance are generally lacking, hindering the study of limiting factors. Estimating such rates is made more difficult by complex population dynamics at non‐breeding sites. In South Carolina, endangered Piping Plovers Charadrius melodus start arriving in July and some depart prior to December (the autumn‐only population) while others remain through at least March (the wintering population). State uncertainty capture‐mark‐recapture models provide a means for estimating vital rates for such co‐occurring populations. We estimated the proportion of the population entering the study area per survey (entry probability) and proportion remaining per survey (persistence rate) for both populations during autumn, and abundance of the wintering population, at four sites in South Carolina in 2006/7 and 2007/8, taking advantage of birds previously colour‐ringed on the breeding grounds. We made fairly precise estimates of entry and persistence rates with small sample sizes. Cumulative entry probability was ~50% by the end of July and reached 95% for both populations by October. Estimated stopover duration for birds in the autumn‐only population was 35 days in year 1 and 42 days in year 2. We estimated a wintering super‐population size of 71 ± 16 se birds in the first year and 75 ± 16 in the second. If ringing programmes on the breeding grounds continue, standardized resighting surveys in the non‐breeding period and mark‐recapture models can provide robust estimates of entry and persistence rates and abundance. Habitat protection intended to benefit non‐breeding Piping Plovers at our coastal sites should be in effect by late summer, as many birds are resident from July to the end of winter.     

The diets of five summer breeding seabirds in Adélie Land,Antarctica

Summary The diets of five breeding seabird species were investigated on Adélie Land in January–February 1982. Stomach contents of Adélie penguins, Pygoscelis adeliae, were sampled by a water off-loading method and of Procellariiformes by spontaneous regurgitation. Diet compositions by mass were: Adélie penguin (79% euphausiid, 18% fish, 3% squid); Cape pigeon, Daption capense, (64% euphausiid, 29% fish, 7% carrion); Antarctic fulmar, Fulmarus glacialoides, (64% euphausiid, 20% carrion, 16% fish); snow petrel, Pagodroma nivea, (95% fish, 2% euphausiid, 1% carrion) and Wilson's stormpetrel, Oceanites oceanicus, (39% fish, 37% euphausiid, 13% carrion, 12% various crustaceans). The present Adélie penguin diet is consistent with those reported in other studies, given our knowledge of geographical variation in food availability. Differences in the diets of fulmarine petrels appear to relate to differences in foraging areas. The snow petrel is a fish-eating bird associated with pack-ice. Cape pigeon and Antarctic fulmar are mainly krill-eaters and we infer segregation along a neritic/oceanic gradient because of the importance of the neritic Euphausia crystallorophias in the former and the oceanic E. superba in the latter.     

The stress hormone corticosterone in a marine top predator reflects short‐term changes in food availability

In many seabird studies, single annual proxies of prey abundance have been used to explain variability in breeding performance, but much more important is probably the timing of prey availability relative to the breeding season when energy demand is at a maximum. Until now, intraseasonal variation in prey availability has been difficult to quantify in seabirds. Using a state‐of‐the‐art ocean drift model of larval cod Gadus morhua, an important constituent of the diet of common guillemots Uria aalge in the southwestern Barents Sea, we were able to show clear, short‐term correlations between food availability and measurements of the stress hormone corticosterone (CORT) in parental guillemots over a 3‐year period (2009–2011). The model allowed the extraction of abundance and size of cod larvae with very high spatial (4 km) and temporal resolutions (1 day) and showed that cod larvae from adjacent northern spawning grounds in Norway were always available near the guillemot breeding colony while those from more distant southerly spawning grounds were less frequent, but larger. The latter arrived in waves whose magnitude and timing, and thus overlap with the guillemot breeding season, varied between years. CORT levels in adult guillemots were lower in birds caught after a week with high frequencies of southern cod larvae. This pattern was restricted to the two years (2009 and 2010) in which southern larvae arrived before the end of the guillemot breeding season. Any such pattern was masked in 2011 by already exceptionally high numbers of cod larvae in the region throughout chick‐rearing period. The findings suggest that CORT levels in breeding birds increase when the arrival of southern sizable larvae does not match the period of peak energy requirements during breeding.     

Crosswise migration by Yellow Warblers,Nearctic‐Neotropical passerine migrants

Yellow Warblers (Setophaga petechia) are abundant breeding birds in North America, but their migratory and non‐breeding biology remain poorly understood. Studies where genetic and isotopic techniques were used identified parallel migration systems and longitudinal segregation among eastern‐ and western‐breeding populations of Yellow Warblers in North America, but these techniques have low spatial resolution. During the 2015 breeding season, we tagged male Yellow Warblers breeding in Maine (N = 10) and Wisconsin (N = 10) with light‐level geolocators to elucidate fine‐scale migratory connectivity within the eastern haplotype of this species and determine fall migration timing, routes, and wintering locations. We recovered seven of 20 geolocators (35%), including four in Maine and three in Wisconsin. The mean duration of fall migration was 49 d with departure from breeding areas in late August and early September and arrival in wintering areas in mid‐October. Most individuals crossed the Gulf of Mexico to Central America before completing the final eastward leg of their migration to northern South America. Yellow Warblers breeding in Maine wintered in north‐central Colombia, west of those breeding in Wisconsin that wintered in Venezuela and the border region between Brazil, Colombia, and Venezuela. Our results provide an example of crosswise migration, where the more easterly breeding population wintered farther west than the more westerly breeding population (and vice versa), a seldom‐documented phenomenon in birds. Our results confirm earlier work demonstrating that the eastern haplotype of northern Yellow Warblers winters in northern South America, and provide novel information about migratory strategies, timing, and wintering locations of birds from two different populations.     

Predation risk and moonlight avoidance in nocturnal seabirds

Unlike most seabird families, the vast majority of small petrel species are nocturnal on their breeding grounds. Further, they reduce markedly their activity when the light level increases. Moonlight avoidance might be a consequence of reduction in foraging profitability, as bioluminescent prey do not come to the sea surface on bright nights. Alternatively, petrels may avoid colonies during moonlit nights because of increased predation risk. We studied predation on petrels by Brown Skuas Catharacta antarctica lönnbergi at Kerguelen, and the influence of moonlight on behaviour of both skuas and petrels, to test the 'predation risk' hypothesis. On the study area, Brown Skuas hunt at night and prey heavily upon the Blue Petrel Halobaena caerulea and the Thin-billed Prion Pachyptila belcheri . Predation risk was higher on moonlit nights, as skuas caught more prey, and particularly more Blue Petrels when the light level increased. Nightly intakes of Blue Petrel and Thin-billed Prion by skuas was related to colony attendance of non-breeders rather than that of breeders. Biometry of prey also suggested that skuas caught a higher proportion of non-breeding birds than was present at the colonies. Predation risk was thus greater in non-breeders and on moonlit nights. Colony attendance by non-breeding Blue Petrels and Thin-billed Prions was also reduced during moonlit nights. Vocal activity, which is mainly by non-breeders, was also drastically reduced when the light level increased in the species suffering the highest predation rate. Our results supported the 'predation risk' hypothesis, although the 'foraging efficiency' and the 'predation risk' hypotheses are not mutually exclusive: the former might explain the moonlight avoidance behaviour of breeding, and the latter that of non-breeding individuals.     

Response of breeding European Storm Petrels Hydrobates pelagicus to habitat change

Mainly through trampling and manuring, ground-nesting seabirds induced significant habitat changes both on vegetation cover and soil in one of the largest French colonies of European Storm Petrel Hydrobates pelagicus, Habitat deterioration led to a high level of erosion and the collapse of many former Rabbit Oryctolagus cuniculus burrows previously occupied by breeding Storm Petrels. The loss of burrows accelerated in recent years since Great Cormorants Phalacrocorax carbo bred on the islet with growing numbers. The main consequence of this disturbance was at first shifting of breeding Storm Petrels from burrows to rocky sites, reflecting some behavioral plasticity to buffer environmental variability. But over 18 years, a significant decrease in breeding numbers of Storm Petrels was recorded and attributed to continuous nest site destruction. Thus, other behavioral responses were also suspected, such as temporary non-breeding or emigration of birds that have to find a new nest site. Such a problem of heavy erosion and loss of nesting habitat could induce serious detrimental effects on burrowing seabirds breeding in a limited number of colonies.