DIET OF HARTLAUB'S GULL LARUS HARTLAUBU AND THE KELP GULL L. DOMINICANUS IN THE SOUTHWESTERN CAPE PROVINCE,SOUTH AFRICA

Steele, W.K. 1992. Diet of Hartlaub's Gull Larus hartlaubii and the Kelp Gull L. dominicanus in the southwestern Cape Province, South Africa. Ostrich 63:68-82.

Hartlaub's Larus hartlaubii and Kelp Gulls L. dominicanus in the southwestern Cape feed on a wide range of prey species. On average, Kelp Gulls feed on larger-sized prey than do the smaller Hartlaub's Gulls. In an undisturbed environment the preferred foraging habitats of Hartlaub's and Kelp Gulls are rocky shores and open, sandy beaches (where sand mussels Donax serra are common) respectively. However, as a result of man's activities, several new foraging habitats have become available including croplands, fishing harbours and refuse dumps. The provision of supplementary food at these new foraging habitats is likely to be the cause of a recent population increase. Kelp Gull chicks are fed predominantly “natural” prey which indicates that the population increase may not be due to enhanced reproductive success, but to improved post-fledging survival of juveniles, which are known to aggregate at sites where supplementary food is provided.     

HARTLAUB'S GULL: EGGS AND INCUBATION

A.J. Williams. 1990. Hartlaub's Gull: eggs and incubation. Ostrich 61:93-96.

At Marcus Island, South Africa in 1979 Hartlaub's Gulls Larus hartlaubii laid eggs between 31 January and 28 February. Clutches were of one to three eggs; and 49% consisted of two eggs. Egg composition, and intra-clutch variation in egg size and weight, are reported. The laying interval was two days, the incubation period 25 days, and the hatching interval 0,8 days.     

POPULATION SIZE,DISTRIBUTION AND DISPERSAL OF KELP GULLS IN THE SOUTHWESTERN CAPE,SOUTH AFRICA

Steele, W.K. & Hockey, P.A.R. 1990. Population size, distribution and dispersal of Kelp Gulls in the southwestern Cape, South Africa. Ostrich 61:97-106.

There are indications that the Kelp Gull Larus dominicanus population of the southwestern Cape Province, South Africa, is increasing. It is proposed that decreased post-fledging mortality due to supplementary food from man's activities, at sites such as fishing harbours and refuse dumps, is likely to be the main cause of this population increase. Kelp Gulls aggregate, and juveniles (first-year birds) are disproportionately abundant, at sites where supplementary food is available. Juvenile dispersal from the two largest breeding colonies, 38 km apart, is in opposite directions.     

BREEDING NOMADISM IN SOUTHERN AFRICAN SEABIRDS-CONSTRAINTS,CAUSES AND CONSERVATION

Crawford, R.J.M., Dyer, B.M. & Brooke, R.K. 1994. Breeding nomadism in southern African seabirds-constraints, causes and conservation. Ostrich 65:231-246.

Fourteen species of seabirds representing three orders and five families breed in southern Africa. Nomadism in the breeding populations of 13 of these seabirds is examined. African Penguins Spheniscus demersus, Cape Gannets Morus capensis. Bank Cormorants Phalacrocorax neglectus, Whitebreasted Cormorants P. carbo and White Pelicans Pelecanus onocrotalus show strong fidelity to specific localities. Caspian Terns Sterna caspia change breeding localities at a low frequency; Cape Cormorants P. capensis, Hartlaub's Gulls Larus hartlaubii and Swift S. bergii and Roseate S. dougallii Terns at a higher frequency. At some localities, Bank Cormorants remain at the same breeding sites for extended periods, whereas Crowned Cormorants P. coronatus, Whitebreasted Cormorants, Kelp Gulls L. dominicanus, Hartlaub's Gulls, Damara Terns S. balaenarum, Swift Terns and Roseate Terns all alter breeding sites. Where space is not limiting, African Penguins may also change sites. Constraints on nomadism include strong attachment to traditional breeding localities, unwillingness to roost at non-breeding localities, and lack of suitable alternative nesting localities. Amongst causes of nomadism are excessive disturbance by humans and South African Fur Seals Arctocephalus pusillus and competition for breeding space, which have resulted in even the least nomadic of the species changing their breeding locations. Regularly nomadic species are likely to respond to environmental cues that enable them either to reduce cost of breeding or to increase reproductive output. Therefore, they are potentially useful indicators of ecological health. Nomadic breeders additionally have the ability to establish new breeding colonies. Some species have used artificial structures to expand their breeding range, and probably also to increase their overall populations. However, nomadic tendencies complicate the conservation of breeding activities. By contrast, breeding species with strong fidelity to traditional localities are more easily protected.     

DISTRIBUTION,POPULATION SIZE AND CONSERVATION OF HARTLAUB'S GULL LARUS HARTLAUBIL

Williams, A. J., Steele, W. K., Cooper, J. & Crawford, R. J. M. 1990. Distribution, population size and conservation of Hartlaub's Gull Lorus hurtlaubii. Ostrich 61: 66–76.

Hartlaub's Gull Larus hartlaubii is endemic to southern Africa, where it breeds between Swakopmund, Namibia and Dyer Island, southwestern Cape Province, South Africa. The species has been re breeding at 48 localities within this range. Between 1984 and 1989 an estimated 12000 pain brered at 31 localities. Twenty-eet percent of the population breeds at Robben Island off the Cape Peninsula, sQuth Africa. Hartlaub's Gull frequently has low breeding success and is considered endangered in Narmbia, where 12% of the poulation occurs. However, the population is increaslng around the urbanmd Cape Peninsula where HartLub's Gull has the potential to become a pest species.     

MIGRATION AND DISPERSAL OF AUDOUIN'S GULL LARUS AUDOUINII FROM THE EBRO DELTA COLONY

Oro, D. & Martinez, A. 1994. Migration and dispersal of Audouin's Gull Larus audouinii from the Ebro Delta colony. Ostrich 65:225-230.

Migratory paths and dispersal patterns of Audouin's Gulls (Larus audouinii) born in the Ebro Delta have been studied, as well as their degree of philopatry. The gulls always disperse southwards from the colony, and only the juveniles seem to disperse in other directions. Dispersal patterns of 2y, 3y and older gulls are similar, wintering mainly on Iberian Mediterranean and Moroccan Atlantic coasts. Juveniles migrate before the rest of the age classes and they winter south of older gulls, in the Senegambia region. However, conclusions on Audouin's Gulls movements are difficult to draw, owing to the large differences in resighting and recovery efforts. Data presented also suggest that Audouin's Gulls do not appear to breed before the age of four years.     

Comparing pellet and stable isotope analyses of nestling Bonelli's Eagle Aquila fasciata diet

Diet analyses are central to the study of avian trophic ecology, and stable isotope analyses have made an increasing contribution in the last two decades. Few isotopic studies have assessed the diet of raptor species, which are more frequently analysed by conventional diet methods such as pellet analysis. In this study, we compare prey consumption estimates of nestling Bonelli's Eagles Aquila fasciata from conventional pellet analysis (in terms of items and biomass) and stable isotopic mixing models (SIAR) using δ¹³C, δ¹⁵N and δ³⁴S of feathers. The pellet analysis showed that European Rabbits Oryctolagus cuniculus, pigeons (mainly Common Wood Pigeons Columba palumbus and Domestic Pigeons Columba livia dom.), Red‐legged Partridges Alectoris rufa, passerines, Yellow‐legged Gulls Larus michahellis and Eurasian Red Squirrels Sciurus vulgaris were the main prey, so they were selected for diet reconstructions in SIAR. At the population level, mean prey consumption estimates were similar for pellets (both items and biomass) and SIAR. At the territory level, the weighted kappa statistic showed good ordinal scale agreement in main prey consumption between items or biomass and SIAR. Although the intraclass correlation coefficient showed poor method agreement when considering all prey in the same analysis, the intraclass correlation coefficients for each prey category showed significant agreement between pellets and SIAR when estimating the consumption of Rabbits, pigeons and Gulls, with lower agreement for passerines and Squirrels. Lastly, there was poor method agreement for estimates of Partridges. Our results suggest an overall agreement between the pellet analysis and SIAR when estimating nestling Bonelli's Eagle diet at both the population and, to a lesser extent, the territory level, supporting the usefulness of isotopic mixing models when identifying the terrestrial and marine components of raptor diets.     

ON THE BREEDING OF THE WHITE-RUMPED SWIFT (APUS CAFFER CAFFER) IN GATOOMA

Crawford, R. J. M., Cooper, J. &; Shelton, P. A. 1982. Distribution, population size, breeding and conservation of the Kelp Gull in southern Africa. Ostrich 53:164:177.

The Kelp Gull Lams dominicanus in Africa occurs coastally between Luanda, Angola and Delagoa Bay, Moçmbique. It breeds between Cape Cross Lagoon, South West Africa/Namibia and Riet River, eastern Cape, South Africa. Censuses of nests and breeding birds at all known southern African breeding localities in the period 1976–1981 indicated that 11 199 pairs bred at 52 localities; 79.5% of this population occurred in South Africa, 57,1% in the Saldanha Bay to Dassen Island region, southwestern Cape. Of the breeding pairs 83% occurred on offshore islands and rocks. Colony size at islands is related to their surface area andMayalso be influenced by food availability and the level of human disturbance. The species breeds in a wide variety of habitats ranging from cliffs and rock stacks to wooden platforms, lowlying vegetation among sand dunes and estuarine sandbars. Any available material is used in the construction of nests, whichMaybe as dense as 4/m² Clutch size is 2–3 eggs. In 1978 breeding took place earlier in South Africa than in South West Africa/Namibia. 92% of the population breeds m sites which are legally protected. Kelp Gulls have decreased or increased in numbers at some breeding localities but there is no clear overall trend. Any increases in colony size near urban areasMayresult in added airstrike hazards.     

Genetic variability in mitochondrial and nuclear genes of Larus dominicanus (Charadriiformes,Laridae) from the Brazilian coast

Several phylogeographic studies of seabirds have documented low genetic diversity that has been attributed to bottleneck events or individual capacity for dispersal. Few studies have been done in seabirds on the Brazilian coast and all have shown low genetic differentiation on a wide geographic scale. The Kelp Gull is a common species with a wide distribution in the Southern Hemisphere. In this study, we used mitochondrial and nuclear markers to examine the genetic variability of Kelp Gull populations on the Brazilian coast and compared this variability with that of sub-Antarctic island populations of this species. Kelp Gulls showed extremely low genetic variability for mitochondrial markers (cytb and ATPase) and high diversity for a nuclear locus (intron 7 of the β-fibrinogen). The intraspecific evolutionary history of Kelp Gulls showed that the variability found in intron 7 of the β-fibrinogen gene was compatible with the variability expected under neutral evolution but suggested an increase in population size during the last 10,000 years. However, none of the markers revealed evidence of a bottleneck population. These findings indicate that the recent origin of Kelp Gulls is the main explanation for their nuclear diversity, although selective pressure on the mtDNA of this species cannot be discarded.     

Prey size selection and cannibalistic behaviour of juvenile barramundi Lates calcarifer

This study assessed the cannibalistic behaviour of juvenile barramundi Lates calcarifer and examined the relationship between prey size selection and energy gain of cannibals. Prey handling time and capture success by cannibals were used to estimate the ratio of energy gain to energy cost in prey selection. Cannibals selected smaller prey despite its capability of ingesting larger prey individuals. In behavioural analysis, prey handling time significantly increased with prey size, but it was not significantly affected by cannibal size. Conversely, capture success significantly decreased with the increase of both prey and cannibal sizes. The profitability indices showed that the smaller prey provides the most energy return for cannibals of all size classes. These results indicate that L. calcarifer cannibals select smaller prey for more profitable return. The behavioural analysis, however, indicates that L. calcarifer cannibals attack prey of all size at a similar rate but ingest smaller prey more often, suggesting that prey size selection is passively orientated rather than at the predator's choice. The increase of prey escape ability and morphological constraint contribute to the reduction of intracohort cannibalism as fish grow larger. This study contributes to the understanding of intracohort cannibalism and development of strategies to reduce fish cannibalistic mortalities.     

DISTRIBUTION,POPULATION SIZE AND CONSERVATION OF THE SWIFT TERN STERNA BERGII IN SOUTHERN AFRICA

Cooper, J., Crawford, R. J. M., Suter, W. & Williams, A. J. 1990. Distribution, population size and conservation of the Swift Tern Sterna bergü in southern Africa. Ostrich 61: 56–65.

The Swift Tern Srerna bergü occurs commonly in coastal southern Africa, with a nonbreeding distribution extending the length of the southern African coast and into southern Angola. The known nonbreeding distribubon of the nominate subspecies S. b. bergü extends from Swakomund, Namibia to Kosi Bay, Natal, South Africa. In Natal its nonbreeding distribution overlaps with that of the subspecies S. b. enigma. The species very rarely occurs inland or far out to sea. The geedin distribution of tre nominate subspecies extends from Swakopmund to Algoa Bay, eastern Cape, South Afcnca.

Twenty-two breeding localities are known in southern Africa, but usually only six to seven are occupied. in any one year. Most breeding localities are marine islands where Swift Terns often breed in muted colonies with Hartlaub's Gulls Lorus hartloubii. In 1984 a complete breeding ansus estimated a breeding population of 4835 pairs. However, 6088 pairs were counted at only three localities in 1988. The size of the annual breeding population varies by a factor of as much as 1,6; although the overall size of the adult population is thought to be reasonably stable with no temporal trends evident. Thirteen breeding localities which are legally proclaimed nature reserves supported 72% of the 1984 breeding population. Existing and perceived threats to the Swift Tern are discussed in detail. To improve further tRe conservation status of the Swift Tern in southern Africa, it is recommended that: unprotected breeding localities be proclaimed reserves, also protecting those not on islands with adequate fences; feral cats be removed from breeding islands where they are present; and fisherman be educated to release entangled terns and not to discard lines and nets.     

Feeding ecology of Kelp Gulls (Larus dominicanus) in marine and limnetic environments

The diet of the Kelp Gull (Larus dominicanus) was investigated at the northern limit of its distribution along the South American Atlantic coast. We used two complementary methods, pellet analysis and stable isotope analysis (SIA), to describe and compare Kelp Gull feeding ecology in freshwater and marine environments. The assimilated diet over two different time scales was investigated via SIA of plasma and red blood cells, blood components with different turnover rates. Fish composed the bulk of the diet of Kelp Gulls in both marine (White Croaker Micropogonias furnieri and Banded Croaker Paralonchurus brasiliensis) and limnetic areas (Armoured Catfish Loricariidae and La Plata Croaker Pachyurus bonariensis), despite the importance of benthic prey from the intertidal zone in samples collected from the marine environment (Wedge Clam Donax hanleyanus and the Yellow Clam Mesodesma mactroides). The fish consumed by the gulls were common discards from fisheries in both environments, and marine bivalves were found at a high density at the marine beach. Diet varied between the different time scales analysed. Conventional diet data generally agreed with stable isotope model estimates, emphasising the importance of using complementary approaches in dietary studies.     

The effects of quantity and quality of prey on population fluctuations in three seabird species

Capsule Population sizes of Common Guillemots Uria aalge, Razorbills Alca torda and Lesser Black‐backed Gulls Larus fuscus were associated with prey abundance but not prey quality.

Aims To examine how the abundance and quality of prey fish affects seabird population size and to test the ‘junk‐food’ or nutritional stress hypothesis.

Methods Analysis of long‐term seabird population size data and Sprat Sprattus sprattus biomass and age‐related weight data using a correlative approach.

Results De‐trended seabird and Sprat population data showed that the abundance of Sprat, the main prey species, was associated with the abundance of seabirds, while no effect of age‐related size of prey on seabird population size was found.

Conclusion As the Sprat population increased so did the seabird populations, regardless of decreases in ‘quality’ of Sprats, implying that more prey fish simply seem to mean more food in this marine ecosystem. No support for the ‘junk‐food’ hypothesis was found and the results contradict suggestions from earlier studies that prey quality is important to top‐predators in the Baltic Sea.     

Intra-population variation in the diet of an avian top predator: generalist and specialist foraging in Great Black-backed Gulls Larus marinus

ABSTRACT

Capsule: Great Black-backed Gulls Larus marinus breeding on Skokholm, UK, fed predominantly on seabirds, rabbits, refuse, and marine prey, with the majority of pairs being dietary generalists, but with some specialist pairs.

Aims: To understand the significance of Great Black-backed Gulls as top predators on a small offshore island with internationally important numbers of breeding seabirds (Skokholm, UK) by quantifying their diet and to determine how this varies within the breeding season, to test for pair-level dietary specialization and to examine the consequences of dietary differences for reproductive performance.

Methods: Regurgitated pellets were collected and analysed from 26 breeding pairs on Skokholm during 2017 and related to breeding success.

Results: Analysis of 1035 pellets revealed that, overall, Great Black-backed Gulls fed on seabirds (48% – mostly Manx Shearwaters Puffinus puffinus), mammals (38% – mostly European Rabbits Oryctolagus cuniculus), anthropogenic waste (7%), and marine prey (7%). Diet varied among pairs with 18 (73%) generalist pairs and 7 (27%) specialist pairs (of which, 5 were bird specialists and 2 were mammal specialists). Diet also varied seasonally, but pair-level dietary diversity was repeatable through the breeding season. Dietary diversity did not covary with breeding success.

Conclusion: Great Black-backed Gulls are top predators on Skokholm. Variation in diet among pairs emphasizes that not all individuals contribute equally in terms of predation. Understanding the incidence of this variation has important ecological implications, particularly where apex predators may exert a strong top-down influence.     

Coevolution of a marine gastropod predator and its dangerous bivalve prey

The fossil record of the interaction between the predatory whelk Sinistrofulgur and its dangerous hard‐shelled bivalve prey Mercenaria in the Plio‐Pleistocene of Florida was examined to evaluate the hypothesis that coevolution was a major driving force shaping the species interaction. Whelks use their shell lip to chip open the shell of their prey, often resulting in breakage to their own shells, as well as to their prey. Mercenaria evolved a larger shell in response to an intensifying level of whelk predation. Reciprocally, an increase in attack success (ratio of successful to unsuccessful attacks) and degree of stereotypy of attack position by the predator suggest reciprocal adaptation by Sinistrofulgur to increase efficiency in exploiting hard‐shelled prey. A decrease in prey effectiveness (ratio of unsuccessful to total whelk predation attempts) and an increase in the minimum boundary of a size refuge from whelk predation for Mercenaria may indicate that predator adaptation has outpaced prey antipredatory adaptation. Evolutionary size increase in Sinistrofulgur most likely occurred in response to prey adaptation to decrease the likelihood of feeding‐induced shell breakage and unsuccessful predation when encounters with damage‐inducing prey occur, coupled with (or reinforced by) an evolutionary response to the whelk's own predators. Predator adaptation to Mercenaria best explains temporal changes in whelk behaviour to decrease performance loss (shell breakage) associated with feeding on hard‐shelled prey; this behavioural change limits attacks on prey to when the whelk's shell lip is thickest and most resistant to breakage. Despite evidence of reciprocal adaptation between predator and prey, the contribution of Mercenaria to Sinistrofulgur evolution is likely only a component of the predator's response to dangerous bivalve prey. This study highlights the importance of understanding the interactions among several species in order to provide the appropriate context to test evolutionary hypotheses about any specific pair of species. © 2003 The Linnean Society of London, Biological Journal of the Linnean Society, 2003, 80 , 409–436.     

6. THE PROBLEM OF NOMENCLATURE

Hockey, P. A. R. 1983. Aspects of the breeding biology of the African Black Oystercatcher. Ostrich 54:26-35. Fifty-five pairs of African Black Oystercatchers Haematopus moquini bred at Marcus Island in 1979–1980. Sixteen pairs laid replacement clutches: the mean interval between loss of the fist clutch and laying of a replacement clutch was 22,2 days. Mean inter-nest distance was 19,4 m. The modal clutch size was 2, with a mean of 1.74. Mean dimensions of 105 eggs were 60,7 × 40,l mm and mean fresh egg mass was 55,s g. There were differences in egg mass and dimensions between eggs in one- and two-egg clutches. Rate of egg loss was high, due mainly to depredation by Kelp Gulls Larus dominicanus promoted by human disturbance. Fledging success was lower at a disturbed site than at undisturbed sites, with highest chick mortality occurring in the first week of life. Mortality of first-year birds of eight days and older was estimated at 48% and 69% in two successive seasons. All juvenile birds dispersed from the natal sites, and were resighted up to 168 km away. Dispersed juveniles were concentrated at the edge of an area of high oystercatcher density.     

Prey Luring Coloration of A Nocturnal Semi‐Aquatic Predator

Body coloration serves a variety of purposes in animals. Diurnal and nocturnal predators such as spiders may use their body coloration to lure prey. We predicted here that the white patches on the forelegs on females of the nocturnal semi‐aquatic spider Dolomedes raptor lure prey, explaining why they are primarily displayed when the spider forages along the water edge. To test our prediction, we developed a color vision model assessing whether the patches are visible to pygmy grasshoppers, the spider's primary prey. We conducted a field experiment using cardboard dummies that resemble D. raptor in size, shape, and color, but with half of them lacking leg patches, and we staged interactions between pygmy grasshoppers and D. raptor with and without leg patches in a greenhouse. We found the white patches to be visible to grasshoppers. The dummies with white patches attracted more grasshopper prey than the dummies without the patches. Moreover, grasshoppers were more attracted to spiders when their white patches were present. Our results supported the hypothesis that the white patches of D. raptor lure prey. Our findings, nevertheless, could not be explained as the spider's body coloration acting as a sensory trap but it should not be ruled out. More studies on a wider range of predators and prey will give more meaningful insights into the co‐evolution of predatory lures and prey sensory modalities.     

SELECTION OF COLONY SITES AND NEST SITES BY COMMON TERNS STERNA HIRUNDO IN OCEAN COUNTY, NEW JERSEY

Common Terns nested on 34 of 259 saltmarsh islands along 46 miles of Ocean County coastline, New Jersey. They nested on low islands of Spartina with less than 12%, windrow ranging in size from 0.6 to 108 acres. All islands selected by terns faced at least two miles of open water from at least one direction. Selection of islands seemed a response to vegetation, size, distance to the nearest island, distance to the nearest shore and exposure to open water. Of the 225 islands without nesting terns, only three fulfilled appropriate criteria. Eighty per cent of nests were situated on windrow. Terns laid larger clutches on windrow compated to those nesting in Spartina. Nearest neighbour distance on islands with low predation varied from 85 to 485 cm and was a function of space, vegetation, and the size of the island. Terns nested closer together on windrow compared with Spartina. Considering all colonies, the mean nearest neighbour distance negatively correlated with the number of nests. Skimmers, Oystercatchers, Laughing Gulls, and Herring Gulls nested in some of the tern colonies. All tern colonies on islands also occupied by nesting Herring Gulls suffered over 10% egg-loss by predation. Proportionally more nests were preyed upon in windrow than in Spartina. Clutch sizes in nests in windrow on islands with predation were significantly lower than those in windrow on islands without predation, even though clutch sizes in Spartina were similar on islands with and without predation. Flooding by exceptional high tides destroyed significantly more nests in Spartina than in windrow. The nests that survived in Spartina were built deeper, and their rim heights were higher than nests destroyed by tides. Choice of colony and nest sites is discussed in terms of the balance of two selection pressures: predation and tidal flooding. Tern nests on windrow are more exposed to predators, but safer from flooding, whereas those nests in Spartina are more susceptible to flooding and less vulnerable to predators. The recent invasion of Herring Gulls into Ocean County has significantly increased the rate of nest predation among affected colonies of Common Terns. It is possible that in future years the terns' behaviour may be modified in response to this new selective pressure.     

Predator size divergence depends on community context

Body size is a multi‐functional trait related to various fitness components, but the relative importance of different selection pressures is seldom resolved. In Carabus japonicus beetles, of which the larvae exclusively prey on earthworms, adult body size is related to the presence/absence of a larger congener and habitat temperature. In sympatry, C. japonicus consistently exhibits smaller body size which is effective for avoiding interspecific mating, but in allopatry, it shows size variation unrelated to temperature. Here, we show that this predator–size variation is attributed to prey–size variation, associated with high phylogenetic diversity in earthworm communities. In allopatry, the predator size was larger where larger prey occurred. Larger adult size may have been selected because larger females produce larger larvae, which can subdue larger prey. Thus, in the absence of a larger congener, variation in prey body size had a pronounced effect on geographic body size divergence in C. japonicus.     

SELECTIVE PREDATION BY TWO SWIFTS AND A SWALLOW IN CENTRAL AMERICA

Samples of flying insects were made at three sites and during dry and wet seasons in Costa Rica and at two sites during the wet season in Panama. Compared with these were the diets of three aerially foraging birds: the Band-rumped Swift Chaetura spinicauda and Rough-winged Swallow Stelgidopteryx ruficollis in Costa Rica and the Short-tailed Swift Chaetura brachyura in Panama. Among insects flying, Diptera accounted for between 52 and 75%, with smaller numbers of Hymenoptera (mean = 15%) and Coleoptera (10%). In bird stomach contents, however, Hymenoptera predominated (59–81%), Coleoptera were less common (mean = 18.5%) and Diptera least common (3.4%) of the three orders. The discrepancy between diets and samples of flying insects could not be accounted for solely on the basis of differential digestion. The composition of the swift and swallow diets was similar; there were no significant seasonal differences in the taxonomic composition of the diet of Chaetura spinicauda. Size distributions of beetle prey fitted log-normal distributions in every case. Mean size of prey was significantly larger than that of insects available in all comparisons. Although there was no seasonal difference in the mean size of prey of Chaetura spinicauda, the variance in prey size, used as a measure of niche width, was significantly larger in the wet than in the dry season. Although insect densities may be higher, foraging time is reduced and the wet season is likely the worse season. This result therefore supports Emlen's hypothesis of narrower niches under optimal conditions and broader niches under suboptimal conditions. For increasing altitude above ground published data show a decrease in the density of insects flying, a decrease in the proportions of taxa favoured by the birds, and a significant decrease in the mean size of beetles. None of these trends can explain the differences between the birds' diets and the samples of flying insects. Preference of prey by these birds is concluded not to be a function of prey size alone, but to involve interaction among prey size, ease of capture, and local density. Because of different flight abilities, insect taxa will differ in ease of capture and, consequently, in mean size preferred by the predator, as well as proportions in the diet. This factor favours Hymenoptera and Coleoptera over Diptera as prey. Local concentrations of winged ants, bees and social wasps, and fig wasps make these prey differentially easy to exploit and explain the preference of Hymenoptera over Coleoptera. Large species of swifts are hypothesized to be the evolutionary consequence of the tendency of aerially foraging birds to exploit local concentrations of single prey species.