From March to July 2000, four seabird colonies in coastal ecosystems of Sinaloa, north-west Mexico were visited: on islets, a sandy beach and a long-abandoned salt-panning flat. There were partial and total breeding failures, most of them due to human activity. Amongst colonies of nine species, the breeding success of three was severely affected. All 250 Royal Tern Sterna maxima eggs were taken for direct consumption; 50% (75) of Blue-footed Booby Sula nebouxii chicks were killed for crab (Portunidae) fishing; and in a Least Tern Sterna antillarum colony (97 pairs), 50% failed due to disturbance from a fishing championship. Some of these species have priority status within Mexican and international conservation regulations. Certain aspects of these problems are discussed, and actions are suggested to balance conservation and the development of economic activities.

Nest-site selection by the sympatric Great Hornbill Buceros bicornis, Wreathed Hornbill Aceros undulatus and Oriental Pied Hornbill Anthracoceros albirostris was investigated in a lowland tropical forest in Arunachal Pradesh, north-east India during 1997–2000. Information on two nests of Rufous-necked Hornbill Aceros nipalensis in higher-elevation forests is also presented. All species nested in live trees of three tree genera, 83% (n = 36) in Tetrameles nudiflora, an emergent deciduous softwood, relatively common in lowland foothill forests. No difference was recorded in nest-tree species or nesting habitats of sympatric hornbills, but there were a few differences in structural characteristics of nest-trees. Cavity size was the main variable separating the three species. Great Hornbills used larger cavities while Oriental Pied Hornbills used smaller cavities closer to riverine areas. Nesting was attempted at 64% of known sites and successful fledging of chicks was 80% overall (n = 72 nests, pooled over 4 years). Nest-trees in disturbed habitats near human habitation were used but were often abandoned or unsuccessful and 50% of all nest-trees were inactive by the end of the study. Potential large nest-trees had a density of 5.9/ha, that of the two most used species was 1.3/ha, and minimum nest densities of all three species was about 1 pair/km2. Interference competition for nest-sites was not observed, despite similarity in nest-tree characteristics, low nest density and high loss of nest-trees. Nest-site availability may naturally limit hornbill populations in the area, and additional loss of nesting habitat to human activities may exacerbate limited availability of breeding sites.

The endemic Seychelles Scops-owl Otus insularis is a Critically Endangered restricted-range species currently recorded only from the montane forest of Mahé, the largest (152 km2) and highest (903 m) island in the granitic Seychelles. Limited research has been conducted on the species and, in particular, details of its breeding biology are poorly known. Behavioural observations were made on 12 pairs by the systematic monthly use of playback of conspecific calls, in conjunction with frequent non-playback territory visits from April 1999 to May 2001. A total of eight nests, including the first nest record, were found on three territories. All were in tree cavities (7–25 m high) and contained either a single egg or chick. This was consistent with additional observations of solitary fledglings (n = 11, from eight territories). Incubation lasted 3–4 weeks and the fledging period was 4–6 weeks (data from two nests). The timing of copulations, in conjunction with the detection of nests and fledglings, suggests that the scops-owl can breed throughout the year with peaks in nesting occurring around May and October. Sex roles during breeding were similar to those of other Strigidae owls: incubation was performed by the female; males courtship-fed the female prior to and during incubation, and the female and chick for the first 2 weeks post-hatching; and both parents fed older chicks and fledglings. Fledglings remained on territory for at least 3 months. Breeding success of study pairs was low: two of eight nests were successful and 11 fledglings (recorded from eight of 12 study territories) were observed in a 26-month period, equivalent to c. 0.5 fledglings per territory per year. Evidence suggests that alien predators may have been a factor limiting breeding success. We discuss the conservation implications of our findings.

Many species of China's Galliformes live in forests and it is often difficult to assess populations of species in these habitats. Such assessments are becoming increasingly important because much of China's forest has been altered through logging and other forms of human activity. After describing and mapping habitat types, we assessed two commonly used methods for counting pheasants (transects and point counts at dawn) in the Xianshennongjia Mountains in the Three Gorges area of Hubei Province in east-central China. Four pheasant species were recorded: Golden Pheasant Chrysolophus pictus, Temminck's Tragopan Tragopan temmincki, Koklass Pucrasia macrolopha and Common (or Ring-necked) Pheasant Phasianus colchicus. Golden Pheasants were detected most often by calls heard during transects and Koklass were recorded mostly during point counts at dawn. Temminck's Tragopan was detected during transects (by sightings) and also by calls at dawn. The relatively few Common Pheasants that were detected were registered mainly during transects. Golden Pheasants were found at the lowest altitudes, closest to human habitation and both Temminck's Tragopan and Koklass lived in various forest types above this. Common Pheasant was found in meadows at higher altitudes. We conclude that the methods used can determine general habitat use, but that no one method is likely to prove sufficient to unravel the detailed pattern of habitat use across all four species. In particular, further study is required to assess the relative importance of different forest types to Golden Pheasant, Temminck's Tragopan and Koklass. It would appear that human impact on the forest has affected the distribution of the pheasant species. For example, Common Pheasant is now absent from low-lying areas and occurs at what appears to be an unusually high altitude in the study area.

The Vietnamese Mekong Delta has undergone much anthropogenic change in the last 100 years, and few areas of natural or semi-natural habitat remain. Despite the likely consequent loss of biodiversity, little documentation exists about the region's avifauna at any time in its history. Here, we present a review of the avifauna of the Vietnamese part of the Mekong delta based on various surveys and other fieldwork carried out since 1988, and an assessment of the few earlier data that were available. A total of 247 species has been recorded from the Vietnamese delta since 1988. Most pre-war data are now considered to be of uncertain validity. Some 50% of the species recorded since 1988 are dependent on wetlands. Of these, 20 are listed as globally threatened or Near Threatened. The delta supports, or is likely to support, internationally important numbers of 21 species of waterbird, including two threatened and three Near Threatened species. Annotations are provided for records of globally or regionally threatened species; for those species for which the Mekong Delta holds internationally important numbers; and those for which records described herein constitute range extensions. Information on the avifauna of the Mekong Delta presented here suggests it is of regional and global importance for the maintenance of wetland biodiversity.

Group size and habitat use of Black-necked Storks Ephippiorhynchus asiaticus were studied over a 30-month period in an unprotected landscape with a mosaic of natural and manmade features in Etawah and Mainpuri districts, Uttar Pradesh, India. Group size was largely one or two and never more than a family of two adults and three young. Group size did not change with season or across habitat types. Black-necked Storks used habitats to different extents across years and seasons. Wetlands were preferred in all seasons and rice-paddies were preferred during the monsoon. The use of irrigation canals was consistent across seasons and was always in proportion to their availability. Irrigation canals and crop-fields, particularly flooded rice-fields, provided some additional habitat to Black-necked Storks, but catered to the population for a very small period of time. While some studies have demonstrated that flooded rice-fields are useful for wintering waterbird species, conversion of natural wetlands to crop-fields will not be advantageous to Black-necked Storks, and these habitats need to be preserved to conserve the species.

This study examines the diets of four hornbill species (Bushy-crested Hornbill Anorrhinus galeritus, Wreathed Hornbill Aceros undulatus, Rhinoceros Hornbill Buceros rhinoceros and Helmeted Hornbill Buceros vigil) common within our 9 km2 study area in the Bukit Barisan Selatan National Park, Sumatra, Indonesia. Line transects and opportunistic sampling were used over one year to collect feeding data, including the tree species and strata in which birds fed, diet items, fruit characteristics and tree crop size. Hornbills fed on 64 species. Figs, non-fig fruits and animals comprised 23.4% (n = 15 species) 51.6% (n = 33 species) and 25% (n = 16 species) of the diet, respectively. Although there was overlap among the diets of the four species, only B. rhinoceros and B. vigil had a statistically significant overlap, primarily due to high fig consumption by both species. B. vigil fed almost exclusively on figs (98.6% of the diet) and a small proportion of animals (1.4%) while 23.1% of the B. rhinoceros diet comprised non-fig fruits and animals (23.1%), with 76.9% figs. A. galeritus and A. undulatus had distinct diets with drupaceous, oily fruits comprising 66.7% and 64.4% of the diets, respectively. We found no significant relationship between the numbers of hornbills visiting fruiting trees and the characteristics of the fruit (weight, length and width). However, maximum crop sizes of diet species significantly influenced the number of feeding hornbills. Hornbill species also differed in the strata in which they fed. A. galeritus and B. rhinoceros were seen in the middle of the canopy 56% and 50.8% of the time, respectively while A. undulatus and B. vigil were observed more frequently in the upper canopy (50% and 74.3% of the time, respectively). The lower part of the canopy was used infrequently by all species. This study suggests that the four hornbill species in our study partition food resources by food type and/or feeding location.

I analyse recent trends in the status and distribution of breeding birds of prey in the Canary Islands (Spain). Populations of Barbary Falcon Falco peregrinus pelegrinoides, Eleonora's Falcon Falco eleonorae, Sparrowhawk Accipiter nisus granti, Buzzard Buteo buteo insularum and Osprey Pandion haliaetus recovered throughout the 1990s. Egyptian Vulture Neophron percnopterus majorensis continues to be seriously threatened, with a total population of only around 130 birds. There is a lack of information on status of nocturnal birds of prey, especially of the Canary subspecies of Barn Owl Tyto alba gracilirostris, whose numbers seem to have decreased. Genetic studies should be carried out to clarify the taxonomic status of the birds of prey of the Canary Islands.

Pink-headed Fruit-dove Ptilinopus porphyreus is a little-known, restricted-range species, endemic to the mountains of the Indonesian islands of Sumatra, Java and Bali. In the period 1981–2002 we conducted surveys throughout its range and compiled data on its presence and absence on 32 isolated mountain complexes. The fruit-dove appeared to be restricted to < 12,000 km2 of forest, scattered over 20 major mountain complexes on the three islands, i.e. three on Sumatra, 16 on Java and one on Bali. It occurred exclusively on mountains of > 2,000 m in altitude, where, depending on the size of the mountain, it reached its lower altitudinal limit between 500 and 1,000 m. The species was found mostly in singles or as pairs but occasionally in flocks of up to 17 individuals. It occurred almost exclusively in forest, feeding on figs and small berries in the upper-canopy. We documented three cases of breeding in the wild. From captive birds it is known that a single egg is laid which is incubated for 20 days, with fledging occurring after another 15–16 days. Based on a study of 104 skins, the breeding season peak in West Java is at the beginning of the dry season. Although Pink-headed Fruit-dove is found in scattered, ever-shrinking forest blocks of mostly < 200 km2, the scarcity of recent field records is more than likely due to its inconspicuous behaviour, and a threatened status is unwarranted as yet.

We used artificial social stimulation (decoys, vocalization playbacks, and artificial nests) to initiate group displays in six (two females, four males) Caribbean Flamingos Phoenicopterus ruber ruber that had not successfully bred since their introduction to Guana Island, British Virgin Islands, in 1992. During a control period prior to the introduction of stimuli, flamingos exhibited no social displays or nest building activities. All flamingos were observed approaching the decoy area as a flock within four hrs of the decoys being introduced, and Head-Flagging displays were exhibited by two birds within the first 24 hrs. In a 12-hr watch conducted two-weeks post decoy introduction, there were significantly more group display behaviours, as well as nest-building, as compared with the control period and immediately after the introduction (3.6% as compared with 0% and 0.35%). Two individuals performed the majority of group displays (although at least one social display posture was observed for each bird) and three birds exhibited nest-building behaviour. Overall, individuals spent most of their time feeding and resting/sleeping (> 95%) during all observation periods. We show for the first time that decoys and vocalization playbacks could have a positive impact on breeding success in the wild by inducing group displays and nesting behaviours in this group of introduced flamingos. We suggest that social attraction techniques may be a useful tool to stimulate breeding in small captive and wild small populations of flamingos.

This paper summarizes current knowledge and outlines future work on the breeding biology of Southern Ground Hornbill Bucorvus leadbeateri in Zimbabwe. All available records since 1900 were analysed, including casual reports by members of BirdLife Zimbabwe and published records. Estimates were made for the start and end of the breeding season, group sizes, clutch size and productivity levels, together with an assessment of preferred habitats and nest-tree species. There is a need for intensive fieldwork to determine aspects of breeding biology such as incubation and nestling periods. Particularly important for sound management and conservation strategies is relative breeding success in different land-tenure systems. The author has started work in a communal area 40 km south of Bulawayo city; land-use systems such as new resettlement areas, commercial farms and protected areas remain to be studied.

Gull-billed Tern Gelochelidon nilotica is classed as Endangered in Europe (Tucker and Heath 1994, Hagemeijer and Blair, 1997), but there have been no detailed studies of the trends in the different populations occurring in Europe and Africa. Here we study the status and trends of the species in Europe and north and north-east Africa. We estimate the total population at 10,500–12,900 breeding pairs, and recognize two biogeographical populations in this region. The western population, comprising colonies in northern Europe (Denmark, Netherlands, Germany), France, Italy, Spain, and north and north-east Africa, consists of at most 6,200 pairs, 1,800 of which are in African colonies. The eastern population, comprising colonies in the Balkan Peninsula, Greece, shores of the Black Sea and Sea of Azov, and Turkey, consists of at most 6,800 pairs. Two trends were observed: a first phase from 1900 up to the mid-1970s in which the northern European populations practically disappeared; and a second phase of stabilization, or even increase, in some of the western colonies, while the eastern population continued to decline. There is a marked concentration of the species in just a few localities in the countries of the Mediterranean basin.

The editor would like to acknowledge the following for acting as referees of papers considered for Volume 14 of Bird Conservation International.

Food habits of Malabar Pied Hornbill Anthracoceros coronatus were studied from December 2000 to December 2001, in the Athikadavu valley, Western Ghats, India. A total of 147 individuals belonging to 18 fleshy-fruited tree species were monitored fortnightly. Thirteen fruit species, including five figs and eight non-figs, were recorded in the birds' diet. The overall number of tree species in fruit and fruiting individuals increased with the onset of summer, the Malabar Pied Hornbill's breeding season. The peak in fruiting is attributed to the peak in fruiting by figs. Figs formed the top three preferred food species throughout the year. During the non-breeding period (May to February), 60% of the diet was figs. During the peak breeding period (March and April), two nests were monitored for 150 hours. Ninety-eight per cent of food deliveries to nest inmates were fruits belonging to six species. Most fruits delivered at the nests constituted figs (75.6%). In addition, figs sustained hornbills during the lean season and should be considered “keystone species” in the riverine forest ecosystem. Two non-fig species are also important. Habitat features and local threats at Athikadavu valley were assessed. The distribution and conservation status of Malabar Pied Hornbill in the Western Ghats was reviewed. Conservation of hornbill habitats, particularly the lowland riparian vegetation, is imperative.

Asian hornbills are primarily frugivorous. We studied the characteristics of fruits consumed by four sympatric hornbill species in Thailand: Great Hornbill (Buceros bicornis), Wreathed Hornbill (Aceros undulatus), Austin's Brown Hornbill (Anorrhinus austeni) and Oriental Pied Hornbill (Anthracoceros albirostris). We compared the frequency of distribution of 11 variables for all fruit species collected in the study area (n = 259) and fruit species consumed by hornbills (n = 73). Our analysis revealed that fruits consumed by hornbills are: (1) large, (2) easily accessible within the canopy, (3) red, purple or black and (4) dehiscent or indehiscent with a thin husk. The range of fruit sizes eaten by hornbills in our study is comparable to that reported from other sites in Southeast Asia and Africa. The large gape width of hornbills enables them to consume large fruits that small frugivores would find difficult to consume.