Article 9 of the Convention on Biological Diversity 1992 requires parties to adopt measures for the ex situ conservation of biodiversity. Within the European Union this has been implemented by the Zoos Directive. The Directive requires zoos and aquariums to adopt a conservation role. Zoos may comply with the Directive by undertaking research from which conservation benefits accrue. However, most current zoo research is concerned with behaviour, environmental enrichment, nutrition and reproduction, and is therefore largely irrelevant to ex situ conservation. It is unlikely that zoos will increase their output of conservation relevant research because most do not have appropriate resources. Furthermore, as an alternative to undertaking research, a zoo may comply with the Directive by engaging in training, information exchange or captive breeding. Most, if not all, zoos already engage in at least one of these activities and therefore may comply with the Directive by doing nothing.

Rees (2005) states in his critical review of the EC Zoos Directive that ‘zoos have no incentive to undertake conservation research because they can legitimately ignore this requirement providing they carry out an alternative conservation measure. Zoos therefore can comply with the EU Zoos Directive by doing nothing.’ Zoo conservation took a major step forward when the Council of EC Environment Ministers agreed in 1998 to an EC Zoos Directive to strengthen the conservation role of zoos. The Directive came into force in 1999 and requires that all Member States set up national systems for the licensing and inspection of zoos. The Zoo Licensing Act 1981 already implements many of the measures in the Directive, including the provision of proper accommodation and care for the animals, keeping up to date records, and taking appropriate measures to prevent escapes. But the requirements that zoos participate in conservation and education activities are new. Although many zoos already participate, the new legislation has made this a statutory requirement. Each European country is now responsible for enforcing the EU Directive with their national zoos. The Directive needs to be general because it applies to animal collections of very different sizes and structures. It cannot be expected that the Directive sets standards higher than is achievable for its smallest members.

In the title of his paper Rees (2005) asked a perfectly valid question with regard to one specific element: the effectiveness of European legislation in enhancing the impact of zoo research in conservation. However, the question was not fully addressed, and the answer may be difficult to determine, due in part to the increasing importance placed on research by zoos independently, and in many cases ahead of, legislation. He discusses at some length the way in which the proportion of zoo based research has shifted in recent years to accommodate and address many of the more technically challenging issues of nutrition, reproduction and population dynamics but without losing sight of the valuable behavioural elements of research on captive animals. Rees manages during the course of his argument to answer many of his own questions, with some clear examples of where and how the practical link can be made between many zoo activities, not just research, and the ultimate delivery of conservation benefits to species. It is worth noting, however, that in many cases the truly measurable and sustainable effect of conservation action on populations, species and habitats requires a timescale that spans generations and therefore may not yet be apparent.

Wehnelt & Wilkinson (2005) suggest that I am confused about the aim of the Zoos Directive, stating that it is ‘… to further the conservation role of zoos rather than increasing research activities.’ On the contrary, they have confused the aim with the means of compliance. Participating in research is but one means by which a zoo may demonstrate a conservation role. I have not suggested that research activity should be increased, or that poorly resourced zoos should develop research programmes, or indeed that research is a compulsory activity. I have merely suggested that much of the research undertaken in zoos is not directly relevant to conservation and, as such, does not fulfil the requirements of the Directive. Wehnelt & Wilkinson have produced no evidence to the contrary.

Headstarting is a management technique employed to enhance recruitment of turtles into diminished or extirpated marine turtle populations. Although there have been numerous projects worldwide, there has been a paucity of detailed investigations into its efficacy. Between 1980 and 2001, 16,422 captive-raised hatchlings and 14,347 yearling green marine turtles Chelonia mydas were released from the Cayman Turtle Farm. Approximately 80% of all turtles released were subject to some form of tagging, including living tags. A total of 392 tagged animals have been recaptured at intervals of up to 19 years. Of this total, 160 individuals were captured in the Cayman Islands and 232 were recorded from other locations within the wider Caribbean and southeastern USA. There was significant variation in the release-recapture intervals at the three countries with most returns (Cayman, Cuba and Nicaragua). A positive relationship exists between time at large and size at recapture and data suggest growth rates comparable to those of wild green turtles in the region. There have been at least six living tag returns, four involving turtles released as yearlings and two involving turtles released as hatchlings. This demonstrates an age at maturity that may be as short as 15–19 years, depending on stage of release. Results show that some headstarted turtles are moving around the Caribbean, surviving for long periods of time, contributing to the local breeding population, and are possibly displaying shifts in habitat utilization with age similar to those recorded by wild individuals.

The Guianas contain one of the largest single remaining tracts of undisturbed tropical rainforest in the world, but this forest and its fauna are facing increased threats. In the north of French Guiana both anthropogenic pressures and conflicts between settlements related to the use of natural resources are growing. Based on surveys in 17 forest sites we show that hunting pressure was the main factor determining current primate species richness, masking the effects of logging or forest type. Three of the larger species, the red howler monkey Alouatta seniculus, black spider monkey Ateles paniscus and tufted capuchin Cebus apella, were less abundant in hunted areas. In the areas around four settlements the harvested biomass of primates was low compared to other game species, but the harvests were close to or beyond the maximal sustainable thresholds for the red howler monkey and tufted capuchin. In French Guiana primates are either fully protected by law (the spider monkey and white-faced saki Pithecia pithecia), or their hunting is restricted to subsistence use (howler monkey, tufted capuchin and wedge-capped capuchin Cebus olivaceus). Most hunted meat is, however, destined for sale. Current conservation policy in French Guiana is limited to legal protection for some species and areas, and laws are poorly enforced. Although large areas of forest and its wildlife are protected simply by their remoteness, there is an increasingly urgent need for the legal protection of all primate species, and the establishment of large protected areas and efficient forest management schemes to minimize the impacts of logging and hunting.

Cabot's tragopan Tragopan caboti, a pheasant that inhabits subtropical montane forests in south-east China, is categorized as Vulnerable on the IUCN Red List. Nesting in trees, it routinely makes use of natural platforms and the old nests of other species, both of which may sometimes be in short supply. This study was designed to test how much use would be made of artificial nest platforms, and to identify factors influencing their occupation. Basketry platforms made of bamboo were erected in parts of Wuyanling National Nature Reserve, Zhejiang Province, China. The tragopans used 16 platforms (8%) in 2002 and 12 (5%) in the following year. A census of the population in spring suggested that a high proportion of the females (41% in 2002, 36% in 2003) in the study area made use of the platforms. There were significant preferences for platforms in mixed conifer/broadleaf forest, as opposed to pure stands, as well as for sites on the upper part of hill slopes. Principal component scores were used to represent variation in 11 variables measured at each platform site. Binary logistic regressions employing these scores as predictors for each year separately did not reveal a consistent model for distinguishing used from unused platforms, although short distances to the forest edge were associated with use in both years. These nest platforms are cheap and easy to erect, and may have the potential to halt or reverse a perceived downward population trend for this species across the whole of its fragmented range. Further trials are therefore advocated.

Despite a ≥75% reduction in the geographic range of Mongolian gazelles Procapra gutturosa over the past 50 years, the species is still recognized as the most numerous large grassland herbivore in Asia. Its actual population size, however, is still disputed, and we therefore estimated its numbers in an 80,000 km2 area in the eastern steppe of Mongolia by driving long-distance (1,200–1,400 km) transect surveys during spring and autumn 2000–2002. Quantitative estimates of gazelle numbers are essential for understanding the causes of changes in the population, and thus devising conservation strategies to assure its long-term viability. Observed herds ranged in size from 1–4,000; among surveys, median herd sizes varied from 14–42. Density estimates varied from 10.7 gazelles km−2 in spring to 11.5 gazelles km−2 in autumn, with total population estimates of 803,820 (483,790–1,330,100 95% confidence interval) and 870,625 (499,432–1,491,278 95% confidence interval), respectively. Confidence limits were wide, and to obtain a coefficient of variation of 20%, transect lengths would need to be extended three- to four-fold. Until more efficient means for conducting population surveys can be implemented, driving long-distance transects, combined with distance analysis, seem to provide the best quantitative estimate of Mongolian gazelle populations.

Major threats to the Vulnerable greater spotted eagle Aquila clanga are habitat loss and hybridization with the lesser spotted eagle Aquila pomarina, but no quantitative studies have been carried out to understand these processes. In Estonia the known remaining A. clanga population comprises only five pure and nine mixed pairs. We describe their habitat at the scale of both the home range and nest site, and use logistic regression models for comparisons with random forest plots and nest sites of A. pomarina. The territories of A. clanga were characteristically near waterbodies, where the species' nesting preference was for dense forest with a pronounced old-growth structure. Habitats of the two species were distinct at both spatial scales. Open natural landscapes were favoured by A. clanga but avoided by A. pomarina. Within its home range A. clanga occupied old wet stands containing black alder, birch and pine, whereas A. pomarina preferred old spruce stands. Multivariate models, which summarized these differences between the two species, did not reveal distinct habitat features of mixed pairs compared with A. clanga pairs. Hence, interbreeding pairs may occupy previous A. clanga territories, and these deserve similar protection to those currently inhabited by A. clanga pairs.

Indonesia's only freshwater dolphin is the facultative Irrawaddy river dolphin Orcaella brevirostris in the Mahakam River in East Kalimantan, Borneo. To clarify the status of this Critically Endangered subpopulation we carried out a series of surveys from early 1999 until mid 2002 on abundance, habitat use, population dynamics and threats. Our estimates of total population size were 33–55 dolphins (95% confidence limits 31–76) based on direct counts, strip-transect analysis, and Petersen and Jolly-Seber mark-recapture analyses of photo-identified individuals. Mean minimum annual birth and mortality rates were similar, i.e. 13.6 and 11.4%, and no changes in abundance >8% were detected over 2.5 years. Dolphins died mainly from entanglement in gill-nets (73% of deaths). The dolphins' main habitat includes confluence areas between the main river and tributaries or lakes. Dolphins make intensive daily use of these confluences, moving up and downstream over an average length of 10 km of river and within a 1.1 km2 area. These areas are also important fishing grounds and subject to intensive motorized boat traffic. Sixty-four percent of deaths (1995–2001) with known location (n = 36) occurred in these areas. Interviews with local residents revealed a generally positive attitude towards the establishment of protected areas for this subpopulation. Because of the dolphins' dependence on areas that are also used intensively by people, primary conservation strategies should be to increase local awareness and introduce alternative fishing techniques.

As most of the pristine forests of South-east Asia have been lost, the ability of its animal species to coexist with humans becomes increasingly important. Dian's tarsier Tarsius dianae, one of the smallest primates, lives in forests of central Sulawesi, Indonesia that are experiencing a dramatic increase in degradation by humans. To evaluate the effects of anthropogenic disturbance on tarsiers we used a comprehensive approach to estimate habitat suitability for these nocturnal insecthunters. On four study plots along a gradient of human land-use we determined population densities, home range sizes, nightly path lengths and group sizes of T. dianae. In total we captured 71 individuals and radio-tracked 30 of these. In more undisturbed sites, population densities were high and travel distances small. We found the smallest home ranges in slightly disturbed forest. In a heavily disturbed plantation densities were low, and ranges and nightly path lengths were large. These results show that undisturbed and slightly degraded forests are the most suitable tarsier habitats, and that focusing on different population parameters could lead to differing conclusions about the suitability of particular habitats.

In the Cook Islands the population of Pteropus tonganus tonganus is thought to be declining, but a lack of knowledge of its status, feeding and roosting requirements has precluded effective conservation plans. We surveyed P. t. tonganus on the Cook Islands through observations, counts and interviews with local residents. We estimated the population to be c. 1,730 on Rarotonga and 78 on Mangaia. A lack of suitable habitat on Mangaia was the most important factor affecting abundance. Overhunting appears to have reduced the populations on both islands. All roost sites were found in undisturbed forest on steep slopes and ridges in the inner and most inaccessible parts of the islands, with roost preference determined by the relative safety from humans rather than food availability. The residents of the Cook Islands seem generally unaware of the serious threat the bats face, with little thought for sustainable hunting. For successful conservation it will be important to alter people's negative perception of these mammals, promoting the value of the bats both ecologically and as a potential source of income from tourists. Habitat protection and enhancement, particularly on Mangaia, will be essential.

The unique primates of south-eastern Madagascar face threats from growing human populations. The country's extant primates already represent only a subset of the taxonomic and ecological diversity existing a few thousand years ago. To prevent further losses remaining taxa must be subjected to effective monitoring programmes that directly inform conservation efforts. We offer a necessary first step: revision of geographic ranges and quantification of habitat area and population size for diurnal and cathemeral (active during both day and night) lemurs. Recent satellite images are used to develop a forest cover geographical information system, and censuses are used to establish range boundaries and develop estimates of population density and size. These assessments are used to identify regions and taxa at risk, and will be a useful baseline for future monitoring of habitat and populations. Precise estimates are impossible for patchily-distributed taxa (especially Hapalemur aureus, H. simus and Varecia variegata variegata); these taxa require more sophisticated modelling.

Morangaya pensilis is an endemic monotypic cactus genus from the southern part of the Baja California Peninsula, Mexico. Previously known only as scattered individuals occurring at a very low density in the highest areas of Sierra de La Laguna, the occurrence of a relatively more dense population (4–7 plants per 250 m2) is reported on Sierra Cacachilas, c. 40 km to the north. Disturbance to the species' habitat includes fragmentation, illegal collection and trade, and land use changes, especially to farming and cattle ranching. Seeds collected from Sierra Cacachilas had a relatively high viability (92%) but recruitment (of 2-month old seedlings) in the area was relatively low, despite above average rainfall in the year of measurement. M. pensilis is eligible for inclusion in Mexico's list of threatened species and fulfils the requirements for categorization as Vulnerable on the IUCN Red List. Management activities that are already taking place in this area include restrictions on cattle and goat grazing and consideration of areas suitable for translocation of the species.

I present data on the decline of Hose's langur Presbytis hosei over a 7-year period in the Kayan Mentarang National Park in the remote northern part of East Kalimantan, Indonesia. In 1996 Hose's langurs were among the most common primates in the Nggeng Bio valley, occurring at densities of over two groups km−2, and could be observed almost daily. A repeat census of the same area in 2003 indicated that these densities had dropped by 50–80%, and observation of the species in the valley had become a rare occurrence. During the 7-year period the forest remained in a relatively pristine condition but, despite being part of a National Park, active protection of the valley was lacking and hunting was common. From interviews with local hunters and Park staff it appeared that hunting for bezoar stones (visceral excretions found in langurs and used in traditional medicine) was the primary reason for the observed decline in Hose's langur. In 1998 a merchant calling at a nearby village had expressed an interest in the stones and guaranteed to purchase them, and this sparked excessive hunting of Hose's langur in the area, to such an extent that 3 years later this hunting was no longer economically viable. This study demonstrates that, with increasing access to markets, hunting large vertebrates for medicinal purposes, even for short periods only, can have a dramatic impact on population numbers. In such cases, habitat protection alone does not guarantee preservation, and more active protection of wildlife is required.