We develop and apply a four-dimensional priority-setting process for the conservation of threatened birds in Venezuela. The axes that we consider are extinction risk, degree of endemicity, taxonomic uniqueness and public appeal. Alhough the first three are relatively objective measures of biological attributes, the last one is a subjective judgement of the likelihood that conservation actions in favour of a species may succeed. By grouping higher priority species according to their geographical distribution within Venezuela, we generate a list of the top priorities to save the country's threatened birds, both species- and bioregion-based. The highest priority species are northern-helmeted curassow Pauxi pauxi, Andean condor Vultur gryphus, red siskin Carduelis cucullata and plain-flanked rail Rallus wetmorei, followed by eight high priority birds, wattled guan Aburria aburri, yellow-shouldered parrot Amazona barbadensis, scissor-tailed hummingbird Hylonympha macrocerca, rusty-faced parrot Hapalopsittaca amazonina, northern screamer Chauna chavaria, torrent duck Merganetta armata, rusty-flanked crake Laterallus levraudi, and military macaw Ara militaris. Northern Venezuela stands out as a significantly higher conservation priority than the south. The Andean Cordillera, Central Coastal Cordillera, Paria Peninsula-Turimiquire Massif Complex, and Sierra de Perijá are the highest priority bioregions, followed by Lara-Falcón Arid Lands and Maracaibo Lake Basin. A final set of combined priorities was determined by integrating all top ranking species and bioregions. Our approach is relatively simple and readily applicable to other taxa and regions.

Depending on the mosaic of habitats that is created, forest fragmentation can reduce the abundance and diversity of forest birds. Temperate deciduous forests in South America are rapidly being replaced by pine plantations, causing changes in habitats for both breeding and migrant birds. We examined differences in avian species richness and abundance in three areas: a reserve with continuous tracts of native forest, forest fragments and pine plantations. Four species were negatively affected by fragmentation, with their abundance declining from continuous forest to pine plantations. Fourteen species were not affected by fragmentation and eight had significantly increased abundance in pine plantations compared to continuous forest. Overall abundance in pooled and ground/understorey birds was higher in forest fragments and pine plantations compared to continuous forest. The abundance of granivorous species was higher in the pine plantations than in the forest reserve, but the abundance of insectivorous species was higher in the forest reserve and in forest fragments than in pine plantations. Thus, forest fragmentation affects birds differentially according to their feeding ecology.

In this study we surveyed all trees ≥10 cm diameter at breast height within 20 0.1 ha plots in a 3,500 ha forest fragment, surrounded by sugar cane fields, of the Brazilian Atlantic forest to compare tree species assemblages at the forest edge (0–100 m into forest) vs forest interior (>200 m). Plots were perpendicular to the margin. The mean number of tree species was significantly higher in the forest interior (35.4 ± SD 7.1 vs 18.4 ± SD 4.4). In addition, forest edge differed from interior in the proportion of shade-tolerant, emergent, large/very large-fruited species, and large-seeded tree species. Among the 134 tree species recorded, 24% were exclusive to the forest edge and 57% to the forest interior. Our results suggest that both the current system of protected areas and archipelagos of small fragments (1) tend to retain only a subset of the original flora, (2) will converge in terms of floristic and ecological composition (biotic simplification and homogenization), and (3) will lose rare and threatened tree species.

Leatherback turtle Dermochelys coriacea nesting has declined on Pacific beaches and as a result the species is considered Critically Endangered. Atlantic populations are, however, also important for the species' survival and therefore we undertook a study to quantify the size and nesting trend of the Caribbean Costa Rica and Panama leatherback turtle rookery. Tag returns show that post-nesting females from the rookery disperse throughout the Caribbean, Gulf of Mexico and North Atlantic. Aerial and track survey results were used to estimate 5,759–12,893 nests per year between San Juan river and Chiriquí beach, making this the fourth largest rookery worldwide. Monitoring results from three beaches (Tortuguero, Pacuare and Gandoca) were used to examine any temporal trend in nesting using nonparametric regression. Nesting appeared to decline slightly from 1995 to 2003 but the trend could be an artefact of interannual variation in nest numbers. Explanations for the difference in nesting trends over the past 15 years for Pacific (rapid decline) and Caribbean (slight decline or stable) rookeries include: (1) hatching success on Caribbean beaches has been higher due to dispersed nesting, (2) fisheries bycatch has been greater in the Pacific, and (3) less overlap between fishing areas and leatherback turtle habitats in the Atlantic. Quantification of human-caused mortality of all life stages and knowledge of the marine habitats used by Atlantic leatherback turtles are required to facilitate the development and implementation of effective strategies to reduce threats and avoid a repeat of the decline that has occurred in the Pacific population.

Fragmentation of forests adversely affects forest-dependent biota, and conservation biologists strive to develop a good understanding of how species respond to changes associated with habitat attrition in order to establish the best conservation strategies. The spatial structure of populations persisting in fragmented landscapes governs their response to habitat fragmentation, and hence dictates the remedial actions that will be most effective for species and habitat conservation. The Taita Hills forests of Kenya are an example of a highly fragmented Afrotropical forest ecosystem embedded in a human-dominated landscape. The spatial structure of the white-starred robin Pogonocichla stellata populations living in indigenous forests across this landscape was examined. Due to its forest dependence and widespread occurrence, the robin was used as a model species to help formulate general conservation guidelines for forest-dependent species and their habitats within this landscape. Results from demographic, genetic and behavioural work point to a mixed spatial structure with elements of patchy population dynamics on a fine scale, and a core-satellite or source-pseudo-sink system on a broader scale. In particular, the findings underscore (1) the importance of dispersal, (2) the importance of small patches, (3) the importance of the largest patch, and (4) the processes underlying problems associated with forest disturbance. We examine the conservation implications of this information, and report on activities already initiated or planned, in line with these findings, for the Taita Hills.

The diversity of amphibians before and after low-level forest exploitation in An'Ala forest in central-eastern Madagascar was compared over the course of 4 years. Neither abundance nor diversity of amphibians generally were significantly affected by low-level selective logging, although the abundance of individual species differed. Mantelline anurans were the most sensitive, in contrast to the tree frogs of the subfamily Boophinae (Mantellidae) and Cophylinae (Microhylidae). The abundance of Mantellinae anurans decreased by 15.8% after logging, whereas Boophinae and Microhylidae anurans increased by 12.1% and 3.7%, respectively. In general, species strongly tied to rainforest habitat showed a non-significant 10.1% decrease in abundance after logging. It appears therefore that amphibians are relatively resilient to a low-level of forest exploitation and their diversity is apparently not affected, at least in the short-term. This and other studies have, however, shown that logging commonly results in a shift in species composition, with species typical of pristine rainforests being replaced by species adapted to secondary habitats.

The Lower Risk/near threatened giant babax Babax waddelli is endemic to southern Tibet. The babaxes, living in groups of 3–8 individuals year-round, favoured edge habitats in forest regions where the group encounter rate during the breeding period was 0.35 groups h−1 and population density 0.04 birds ha−1, and in scrub regions they depended on the patches covered with higher bushes in which the group encounter rate and population density were 0.32 groups h−1 and 0.03 birds ha−1 respectively. The global population of the babaxes was estimated to be 120,000 birds. Reproductive and behavioural data were collected from scrub environments in a valley. Egg laying occurred between early May and early June, and clutch size ranged from 2 to 4 eggs, 2.9 on average. Incubation and fledging lasted 16–18 days respectively. Nesting success, measured as the proportion of clutches from which at least one young survives to fledging, was 75%. Some of the babax's nests were cooperatively attended by group members and the others only by both parents. The most preferable nesting areas were in dense bushes more than 1.8 m in height (mainly Rosa sericea, Berberis hemleyana, Cotoneaster divaricatus, Sibiraca angustata and Salix sclerophylla). Such strong dependence on high bushes leaves babaxes more susceptible to vegetation degradation. As a result, the increasing clearance of scrub habitats for firewood is a major current threat to the birds around the mid-Yalong Zangbo River, a region with the densest human population and most rapidly developing economy in Tibet. Their naturally low population density and strongly social grouping behaviour present extra risk factors that have to be taken into account. The importance of vegetation protection for long-term survival of this endemic species is highlighted.

We conducted a survey in the western Himalaya of India to assess animal extraction patterns. Data on animal species and their extraction patterns, their importance to the respondents, and reasons and methods of hunting were collected using structured questionnaires. Twenty-three species of large mammals and Galliformes were present in the area, 18 of which were hunted around at least one village. Of special concern were several threatened species that were hunted around most villages were they occurred, although the impact of removal on wild populations is not clear. The main reason for hunting was to supplement animal protein, although some animals were also killed for sale of meat and their parts. The establishment of community-managed forests has not had an impact on extraction rates. Assessment of the impact of hunting on the threatened species in particular is urgently required.

Sea moths (family Pegasidae) are little-studied benthic fish, found throughout the Indo-Pacific. Two species of sea moths, Pegasus volitans and Eurypegasus draconis, are caught incidentally in illegal trawl gear in the Philippines and sold into the dried fish trade. Approximately 130,000–620,000 P. volitans and 130,000 E. draconis were landed off north-western Bohol alone in 1996. An additional 43,000–62,000 sea moths (predominantly P. volitans) were caught live for the aquarium trade. Catch per unit effort for P. volitans was double that of E. draconis, probably because of its occurrence in shallower waters where fishing effort was concentrated. Sea moths may be unsuited for heavy exploitation as they occur at low densities. Moreover, a female-biased catch could lower the effective population size, given the reported monogamy amongst sea moths. No population data were available for a complete conservation assessment, although divers surveyed did report declines in their abundance.

Faunal surveys in Kalimantan have been biased towards primates in protected forests close to the coast. Relatively little has been documented on other animal species, particularly in the vast interior forests. The results of a 1996–97 census of nine large mammal and eight hornbill species in tropical lowland forest in Barito Ulu, Central Kalimantan are reported here. Pigs Sus barbatus had the highest biomass, but this was due to large numbers migrating through the study area over 4 months and the resident population is probably low. Langurs Presbytis rubicunda and hybrid gibbons Hylobates mulleri × agilis had the highest biomass of all resident species. Orang-utans Pongo pygmaeus were absent from the area during the study period and pig-tailed macaques Macaca nemestrina were rarely seen. The resident hornbill species (Anthracoceros malayanus, Anorrhinus galeritus, Buceros vigil and B. rhinoceros) had high densities compared to that reported from lowland areas, but overall hornbill density was low due to the absence of the nomadic Aceros corrugatus and A. undulatus, except during peak fruit abundance. Sun bears Helarctos malayanus, long-tailed macaques M. fascicularis, muntjacs Muntiacus spp. and mouse deer Tragulus spp. were at low densities. Density of two large squirrel species, Ratufa affinis and Sundasciurus hippiurus, was lower than has been reported in Sarawak, but the density of Prevost's squirrel Callosciurus prevostii was higher. We discuss hunting pressure, isolation, low abundance of large fruit trees, poor soils, and specific habitat preferences as possible explanations for the low mammal and hornbill density at Barito Ulu.

The white-winged guan Penelope albipennis (Aves, Cracidae) is a Critically Endangered bird endemic to northern Peru. We surveyed a protected area suitable for its reintroduction north of the species' current range. Following IUCN guidelines, four factors were evaluated: diversity and quantity of plant species that are part of the white-winged guan's diet, year-round water sources, resting and roosting cover, and extent of undisturbed forest. Several suitable small valleys or quebradas were inspected, but our evaluation focused on three. We identified all shrub and tree species to compare with habitat currently used by the guan. Cover for nesting places was assessed visually. We found species composition and microhabitat types to be similar to habitat presently used by wild guans. Thus, all habitat components are favourable for reintroduction. The high degree of protection of the area makes it particularly suitable as a reintroduction site. We recommend experimental extension of the species current range to mitigate the risk of extinction posed by catastrophic events.

The Hainan black-crested gibbon Nomascus sp. cf. nasutus hainanus is on the verge of extinction. Following studies in the late 1980s and early 1990s only limited observations were made of the remaining population in the Bawangling National Nature Reserve on Hainan Island, China, and the most optimistic estimation of the population size was 23 individuals. To determine the present status of this gibbon, we surveyed the western part of the Reserve, from where the gibbons were previously known, using transects (November–December 2001), listening post (February–April 2002) and interviews with reserve staff (November–December 2001 and February–April 2002). Three groups of gibbons were sighted comprising at least 10–12 individuals including juveniles and infants. A fourth group, possibly comprising 2–7 individuals, was heard but not seen. Compared to an unpublished survey by reserve staff in 2000, the population appears to have decreased. Surveying of this gibbon population will continue, in particular to determine the composition of the one group that was heard but not seen and to carry out a survey of the eastern part of the Reserve to determine whether or not any gibbons survive there. The Forestry Bureau of Hainan is applying to the Central Government for an enlargement of the area of the reserve and more staff to help prevent illegal poaching and logging.

The Fijian island of Taveuni is being proposed as a potential World Heritage Site because much of its flora and fauna, including many endemic species, remains intact. The greatest threat facing Taveuni's wildlife is a potential incursion by the invasive small Indian mongoose Herpestes javanicus from nearby islands, and anecdotal reports have suggested that the mongoose may already be present. To determine if mongooses were present on Taveuni 40 traps were set in two locations close to the main ports of entry, in typical mongoose habitat. Normally mongooses are easily caught if they are present, but in 204 trap nights none were captured, although eight feral cats and 11 rats were caught. None of the 31 people questioned from around the island had seen the mongoose. A list of seven recommendations is provided for a management strategy to prevent any possible incursion by mongooses and to eradicate any arrivals at the earliest opportunity.