Mountain gazelles were reintroduced to central Arabia during 1991–95. Hawtah reserve was searched for gazelles during the 1998–99 winter. Gazelles were seen in one wadi system and their signs were found in several others and on the plateau. Sightings were used to calculate the minimum number of gazelles in the Matham wadi system, which previously held most of the population. During October–November 1998, the minimum number was 64 per cent less than 4 years earlier. Frequent observation of recent signs in areas where no gazelles were seen suggested that daytime sightings alone were no longer adequate for monitoring this population. The decline in the number of gazelles seen, an increase in their flight distance and an apparent change in their activity patterns were consistent with the rangers' claim that poaching had commonly occurred. Poaching started after reserve management built, without adequate consultation, a new fence that was intended to bar local people from part of the reserve. Management lessons include the need for the following: continued monitoring of reintroduced populations after the initial postrelease phase; long-term dialogue with local people; effective law enforcement; and the management of aridland domestic livestock in ways that prevent inter specific competition for food causing the elimination of wild ungulates.

Abstract The Critically Endangered Antiguan racer Alsophis antiguae is confined to Great Bird Island, a 9.9 ha (24.5-acre) islet off the north-east coast of Antigua in the Lesser Antilles. This island represents well under 0.1 per cent of the species's historical distribution range. During the past 5 years, the total number of racers aged 1 year or more has fluctuated between 51 and 114, and currently stands at approximately 80. Since 1995, the Antiguan Racer Conservation Project (ARCP) has en-deavoured to save this harmless snake from extinction by using a combination of education, conservation breeding, habitat restoration, local capacity building and applied research. The Antiguan racer's ecology and population dynamics have become well understood after 5 years of intensive study, and the species has evidently benefited from the project's rat eradication programme. The snakes are still seriously threatened by other intrinsic and extrinsic factors, however, including inbreeding depression, frequent hurricanes, invasive predators and deliberate killing by tourists, as well as the problem that Great Bird Island is too small to support more than about 100 individuals. This paper describes the activities and impact of this project to date, and outlines a series of conservation activities to safeguard the long-term future of the species, which include reintroduction of the Antiguan racer to restored islands within its former distribution range.

Birds restricted to islands are susceptible to extinction, and burrow or ground-nesting birds are particularly vulnerable to introduced mammalian predators. Human intervention has also played a vital part. Birds have been used as a source of food, and in more recent times the rarer species have suffered from specimen and egg collection. The island of Madeira and its resident species, which include the endemic Zino's petrel or Madeira freira Pterodroma madeira, are no exception. From subfossil evidence, this bird was once abundant. It was first recorded in 1903, and was already limited to the high central mountain massif of Madeira. By the middle of the century it was considered extinct, but a relict population was rediscovered in 1969. By 1985, all known breeding attempts were disrupted by introduced rats, to the extent that no young fledged. In 1986 the Freira Conservation Project was founded with the aim of increasing the population of Zino's petrel, by controlling rats and human interference, the principal perceived threats to the species. This control was extended to cats after the disaster of 1991, in which a cat(s) managed to get onto one of the breeding ledges and kill 10 adult birds. The results of these efforts have been positive and the small colony is making a slow, but steady recovery. To maintain this success, a conservation strategy for the future is suggested.

This paper evaluates the ecological consequences of hunting by comparing mammalian densities, biomass, relative energy consumption and community structure between sites with different levels of hunting pressure. Hunting is carried out mainly by colonists who farm on the edge of Atlantic forest fragments in the State of São Paulo, Brazil. Mammals were studied over a period of 18 months, along 2287 km of line transects. Transects were distributed among two protected sites, one slightly hunted site and two heavily hunted sites. Tapirs, the two peccary species, brocket deer, armadillos and agoutis are preferred by hunters in the region. Primates are not hunted in the region. Hunting has affected community structure, with ungulates dominating mammalian biomass at protected sites and primates dominating at hunted sites. This has caused an ecological inversion in the hunted areas of the Atlantic forests. In Amazonian regions of the Neotropics hunting is more evenly distributed among primates, large rodents, and ungulates and has resulted in an opposite inversion, with hunted sites having lower primate biomass. Atlantic forests are very susceptible to the possible ecological imbalances induced by hunting by humans, and this must be considered for management and conservation programmes.

Large populations of marine turtles breeding in the Cayman Islands were drastically reduced in the early 1800s. However, marine turtle nesting still occurs in the islands. The present-day status of this nesting population provides insight into the conservation of marine turtles, a long-lived species. In 1998 and 1999, the first systematic survey of marine turtle nesting in the Cayman Islands found 38 nests on 22 beaches scattered through the three islands. Three species were found: the green Chelonia mydas, hawksbill Eretmochelys imbricata and loggerhead Caretta caretta turtles. Comparison with other rookeries suggests that the small number of sexually mature adults surviving Cayman's huge perturbations may be impeding population recovery. This shows the need to implement conservation measures prior to massive reductions in population size.

The results of a field census of Alaotran gentle lemurs Hapalemur griseus alaotrensis carried out in February and March 1999 are reported. The results are compared with a census carried out 5 years earlier in 1994. Both censuses followed the same methods and were carried out by the same team, using direct observation from canoe. Mean group encounter rates for the lemur were assessed in each location, which allowed us to calculate the relative group density in each site. Qualitative information on habitat condition (on the basis of plant diversity, vegetation height and evidence of burning) was gathered at each site. Additional information, mainly on lemur hunting, was acquired through interviews with local people. In most sites group encounter rate was at least 50 per cent lower than 5 years before. Taking into account unusually low water levels because of a drought at Lac Alaotra in 1999, we estimate that these encounter rates reflect a 30 per cent decline in the total population size over the last 5 years. The main cause of this dramatic decrease was humaninduced fire (habitat degradation) coupled with heavy hunting (poaching of lemurs for food). The importance of assessing regularly both population status and threats, and of adjusting conservation actions accordingly is emphasized.

Rodents account for 40 per cent of living mammal species. Nevertheless, despite an increased interest in biodiversity conservation and their high species richness, Rodentia are often neglected by conservationists. We attempt for the first time a world-wide evaluation of rodent conservation priorities at the genus level. Given the low popularity of the order, we considered it desirable to discuss identified priorities within the framework of established biodiversity priority areas of the world. Two families and 62 genera are recognized as threatened. Our analyses highlight the Philippines, New Guinea, Sulawesi, the Caribbean, China temperate forests and the Atlantic Forest of south-eastern Brazil as the most important (for their high number of genera) ‘threat-spots’ for rodent conservation. A few regions, mainly drylands, are singled out as important areas for rodent conservation but are not generally recognized in global biodiversity assessments. These are the remaining forests of Togo, extreme ‘western Sahel’, the Turanian and Mongolian-Manchurian steppes and the desert of the Horn of Africa. Resources for conservation must be allocated first to recognized threat spots and to those restricted-range genera which may depend on species-specific strategies for their survival.

Eight lesser anteaters Tamandua tetradactyla rescued from the rising waters of a dam at Serra da Mesa, Minaçu, Goiás, Brazil, were tranlocated to other areas and tracked using radio telemetry for periods of up to 10 months, from December 1996 through February 1998. With the exception of one, or perhaps two, female(s) that left the area while the radio collars were in place, the anteaters stayed within 2.17 km of their release sites and appeared to thrive. The results suggest that it is feasible to translocate small numbers of anteaters into new areas of suitable habitats without adverse effects on resident anteaters.