Endangered southern river otters Lontra provocax (huillin) are confined to southern Chile and three isolated areas of Argentina, the most important being the Nahuel Huapi National Park, where distribution surveys were conducted in 1983, 1995 and 2000. We used these survey data together with a Geographical Information System analysis to elucidate the factors that are important in determining huillin distribution at different ecological scales. There were significant heterogeneities in the distribution of huillin between river basins, between habitats, between lakes relative to dispersal routes and topography, and between sample points within a lake. From these analyses three factors emerge as potentially related to huillin distribution, and further elucidation of these are suggested as priorities for research to underpin the species' conservation. These factors are: a) the distribution of crustacean prey, with absence of huillin from one river basin and from mountainous streams correlated with low densities of crustaceans in these places; b) human habitation, which may explain why huillin dispersed southwards, but not northwards, from the main lake, although underwater visibility may also be relevant; c) the presence of the introduced American mink Mustela vison, because although there was a positive relationship between the occurrence of mink and huillin in Lake Nahuel Huapi, there was a negative relationship in their occurrence between other lakes. We discuss possible explanations for all three factors.

The distribution of the Endangered southern river otter Lontra provocax population in southern Chile has contracted, causing concern among otter specialists and conservationists. The main factor contributing to this decline is the loss of suitable habitat. Because the Chilean Irrigation Law promotes land drainage, the removal of riparian vegetation and canalization of lowland rivers in several provinces, this situation could deteriorate further. Using an Index of Revisitation Rate (IRR), we assessed the influence of woody debris, riparian vegetation, river bank morphology and river canalization on the use of rivers and streams by the southern river otter. Surveys were conducted on foot, on a monthly basis, between April 1999 and March 2001 in the south of Chile. The surveys covered 13 sections of rivers and streams, 3 km long and <50 m wide. River and stream banks with high density of riparian vegetation, woody debris and exposed roots had the highest IRR; most of the highly marked sprainting sites were found in this habitat. Canalized rivers and streams lacking these characteristics had the lowest IRR and hardly any highly marked sprainting sites. The data indicates that otters can tolerate some degree of riparian vegetation removal, provided there is minimum use of floodplains by humans. Temperate evergreen ombrophilus swamp forests and shallow watercourses with woody debris or gravel, riparian vegetation and a plentiful supply of macrocrustaceans appear to be the preferred habitat and conditions for the otter and may be key to its survival in freshwater habitats.

Interview surveys of 116 coastal residents and fishermen and surveys by motorboat were conducted in 2000 and 2001 in French Guiana in order to assess the distribution, habitats and conservation status of the West Indian manatee Trichechus manatus. There were 72 reported sightings of manatees, and we made 10 sightings during river surveys. Interviews suggested that manatees may be less abundant than in the recent past, but are still present and regularly sighted all along the coast and in estuaries up to 80 km inland. The main reason for any recent decline is probably hunting for meat, pressures from development and, in some areas, pollution from maritime traffic. Although only opportunistically caught, manatees are still poached. In some areas they are killed for cultural beliefs, but only on a small scale. The main habitat for the manatee in French Guiana is estuarine mangroves. A survey at Coswine, an area where manatees have been sighted year-round, indicated that this mangrove area is characterized by biotic and abiotic factors within the range known to provide suitable habitat for the species. However, habitats for manatees in French Guiana are limited and receive little formal protection, and the species' status is based more on fortuitous protection than on any active conservation. Public awareness programmes for the manatee are currently being developed by local NGOs, but a regional conservation programme is also required.

During 1998–2000 extremely low densities of gorillas Gorilla gorilla gorilla and chimpanzees Pan troglodytes troglodytes were found in the Minkebe Forest block in north-eastern Gabon. When compared with data collected before 1994, these data suggest a catastrophic decline in ape populations in the area. We believe that this decline was caused by a disease epidemic. The period of decline corresponds with the Ebola outbreaks of 1994 and 1996 that occurred in the human population in the same area. Deaths of gorillas and chimpanzees were associated with both Ebola outbreaks. Data from nearby sites indicate that the epidemic was limited to the Minkebe Forest. Occurrence of such epidemic die-offs should be taken into account in conservation strategies for the long-term survival of ape populations. At the time of writing, an Ebola epidemic among humans in the Zadié Department east of Minkebe Forest has resulted in 53 deaths. In the neighbouring Republic of Congo, authorities have reported 43 deaths and at least 12 other cases of Ebola. These epidemics are believed to be linked to the handling and eating of dead apes.

We report the results of a survey of Fischer's turaco Tauraco fischeri in Unguja (otherwise known as Zanzibar). We estimated the species' range, population size and habitat selection using unlimited distance point counts. The main results are: 1) Fischer's turaco is more widely distributed in the coral rag zone than previously thought, with an area of occupancy estimated at 56 km2, 2) the species does not occur in agricultural habitats and tree plantations, 3) within natural habitats it selects places with denser tree cover, 4) food availability does not differ between occupied and unoccupied sites, and 5) the turaco population on Unguja was estimated at c. 1,400 individuals. Only 44% of the population is in protected areas, and habitat degradation is occurring at a high rate. Only 16% of the habitat occupied by Fischer's turaco on the island has low signs of human impact. Present levels of exploitation of natural habitats on Unguja appear to be unsustainable and it seems likely that all unprotected forests on the island will be lost within a few years. The range of Fischer's turaco is highly fragmented and the population is likely to be decreasing. This turaco was often observed in places where other threatened species were also present, suggesting that it could be both a flagship and an umbrella species for the implementation of wider habitat conservation strategies. The conservation status of Fischer's turaco should be reconsidered, with its IUCN threat-category upgraded from Lower Concern: near threatened to Vulnerable.

Olive ridley turtles Lepidochelys olivacea nest along the east and west coasts of India, with major mass nesting beaches in the state of Orissa. The coast of Andhra Pradesh, the state immediately south of Orissa, has sporadic nesting of olive ridley turtles and is believed to form part of the migratory route of the turtles that nest in Orissa. A survey of nesting beaches and offshore waters of the Andhra Pradesh coast was carried out from November 2000 to April 2001. Preliminary interviews and secondary data were used to determine potential nesting beaches. During January–March 2001 intensive surveys of seven beaches and monthly surveys of the rest of the coast provided a lower bound of c. 4,000 nests along the Andhra Pradesh coast. Nesting densities were higher at beaches near river mouths, at 60–100 nests km−1 in northern and central Andhra Pradesh, and 15–20 nests km−1 in southern Andhra Pradesh. Sightings and incidental catch in experimental trawls indicated the presence of olive ridley turtles in offshore waters. Fisheries related mortality is the major threat to the species, with nearly 1,000 dead turtles being washed ashore during January–March, but depredation of eggs by humans and feral animals was also widespread. Conservation efforts need to address these issues, and also the effects of coastal development and artificial illumination, especially at beaches that support relatively high densities of nesting olive ridley turtles.

The Critically Endangered Burmese star tortoise Geochelone platynota is endemic to the dry zone of central Myanmar. Although quantitative assessments are lacking, populations are believed to be in decline throughout its limited range due to over-collecting for international markets. We conducted star tortoise surveys during 2000–2001 at two sites in central Myanmar: Minzontaung Wildlife Sanctuary, a protected area established in 1998–1999, and Myaleik Taung, an area recently proposed for protected status. G. platynota populations of global conservation significance were found at both sites. Because tortoises are protected by local religious beliefs and commercial collecting is so far non-existent, the Sanctuary and Myaleik Taung offer excellent prospects for star tortoise conservation. The Sanctuary is well patrolled, protected from wildfires, and access is rigorously controlled. Recommended conservation actions at Myaleik Taung include management of livestock grazing, prescribed burning, and elimination of fuelwood cutting.

We present the first population estimate for the Natuna leaf monkey Presbytis natunae, endemic to the island of Bunguran, Indonesia, based on a 2-month survey. Bunguran has a land area of only 1,605 km2 and was until 1980 largely covered in primary forest. At present primary forest exists only in small patches within a matrix of logged forest, which covers nearly 70% of the island. Natuna leaf monkeys are confined to forest and have a preference for primary lowland forest patches. Groups average 3.5 ± SD 2.0 individuals and occur in a density of 2.3 ± SD 1.1 groups per km2. Extrapolation of the density estimates to the entire island indicates that <10,000 individuals remain in two subpopulations. On Bunguran these monkeys make popular pets and the species is threatened in part by the demand for captive individuals but mostly by continuing degradation and reduction of forest. An assessment of the species following IUCN threat criteria indicates that it should be categorized as Vulnerable. We make a number of recommendations for the conservation of the species, most importantly the implementation by the district government of Natuna of two strict conservation areas on the island, and the maintaining of a larger area of sustainably used natural forest.

The Makgadikgadi Salt Pans in Botswana are one of the most important breeding sites in southern Africa for lesser flamingos Phoeniconaias minor and greater flamingos Phoenicopterus ruber roseus. Much of flamingo migration behaviour is unknown and there has been speculation on the pattern of flamingo movements to and from Makgadikgadi and their dispersal throughout southern Africa. We carried out the first satellite tracking of flamingos in southern Africa to find out where lesser and greater flamingos go after leaving Makgadikgadi. In July 2001 five lesser and three greater flamingos were tagged. Following migration from the pans, one of the greater flamingos flew west to the coast of Namibia, the other south to a small wetland in South Africa. The lesser flamingos moved south-east from Makgadikgadi to South Africa and Mozambique. Movement by both species was nocturnal. This work shows that flamingos migrate from all over southern Africa to Makgadikgadi to breed. It also shows that, during the non-breeding season, movement is widely dispersed and nomadic among a network of wetlands around the subcontinent. Small wetlands, often unrecognized as important for conservation, provide valuable feeding sites and migration staging posts along flamingo migration routes. This highlights the need for the conservation of the network of small wetlands around southern Africa, which are often under threat from anthropogenic activities, to protect two high profile bird species in decline.

Populations of European polecat Mustela putorius have suffered a significant decline over the last few decades, although this trend may be reversing in some regions. In Mediterranean areas, which are generally at the edge of the species' range, there is little information on the distribution and conservation requirements of this species. This study analyses the relationship between habitat types and polecat distribution in the Valencia region of eastern Spain. Validated responses to questionnaires sent to naturalists, forestry agents, hunters and other groups likely to have seen polecats were used to record the presence of polecats in 10 * 10 km squares. Polecats were recorded only in areas of low human population density located at high altitudes, in contrast to the more well-studied populations of polecats elsewhere in Europe, where the species is associated more with human-modified environments. Montane pine forests appear to be the preferred habitat type for polecats in eastern Spain, and their absence from the extensive areas of low-lying scrubland, which are produced by a continuous history of fire disturbance, may be due to the relatively low diversity and abundance of potential prey in this habitat.

The onager Equus hemionus onager, a wild ass endemic to Iran, is categorized as Critically Endangered on the IUCN Red List. Its biology and conservation requirements are poorly documented. We report our observations, made in 1997 and 2000, on the behaviour and ecology of the two remaining populations, located in the Touran Protected Area and the Bahram-e-Goor Reserve. Recent population counts by the Department of Environment of Iran (471 in the Protected Area and 96 in the Reserve) are markedly lower than the estimate of 600–770 made in the 1970s in the Touran Protected Area. We observed social interactions between stallions and mares outside the breeding season that contrasts with the known social structure of this subspecies. Poaching, competition with domestic animals, removal of shrubs for domestic use, and land conversion have been identified as the main threats to the two remaining onager populations. In addition, geographical isolation could cause the loss of genetic variability in these two relatively small populations, and also makes them more susceptible to the potential effects of stochastic events such as drought or disease. Public awareness, appropriate protection, and scientific studies must be urgently supported by both national and international organizations in order to prevent the extinction of these two apparently dwindling populations of onager.