have developed a flexible package of training materials to help build the capacity of conservationists to address the social and rights aspects of conservation. The project, known as INTRINSIC, was funded by the Cambridge Conservation Initiative Collaborative Fund. Materials were tested with students on the Cambridge University Masters in Conservation Leadership and with conservation practitioners in East Africa. Biodiversity conservation is essentially a social process, involving as it does decisions about access to, and the use, values and protection of nature. As such, conservation inevitably entails both social costs and benefits, and the social context is likely to affect the efficiency and effectiveness of any conservation initiative. Unfortunately conservation practitioners often lack the knowledge and skills to address and integrate rights and other social issues into their work. The typical pathway to a professional conservation career involves the pursuit of educational opportunities and formal qualifications in natural sciences. However, the contemporary practice of conservation necessitates working with people, local communities and groups, and requires knowledge and skills about social systems that are often not gained through these traditional academic pathways. A number of authors have pointed out discrepancies between conservation course content and the skills needed, and have bemoaned the lack of training in the social dimensions of conservation (e.g. Saberwal & Kothari, , Conservation Biology, , -; Jacobsen & McDuff, , Conservation Biology, , -; Fisher et al., , Oryx, , -). The INTRINSIC training package aims to help address this capacity gap and improve conservation policy and practice by increasing environmental and social sustainability, thereby enabling positive, equitable outcomes for both nature and people. The INTRINSIC materials, which can be found at http:// bit.ly/IntrinsicManual, comprise a trainers' guide and accompanying set of slide presentations designed to be customized for the particular context in which the training is to take place. It is expected that users will have some experience of working in conservation but in-depth knowledge of the specific social issues covered in the guide is not required. The developers envisage that delivery of the training can be a learning opportunity for trainers as well as participants. Subjects covered include community and social diversity, gender, conflict management, livelihoods and

Protected area (PA) extent has increased significantly over the last 150 years globally, but it is yet unclear whether progress in expanding coverage has been accompanied by improved performance in ecological representation. Here, we explore temporal trends in the performance of PA networks in representing > 16,000 vertebrate and plant species in tropical Andean countries based on species bioclimatic niche modelling. We use a randomization analysis to assess whether representation gains over time (1937–2015) are the expected consequence of increasing the overall area of the network or the result of better designed networks. We also explore the impact of climate change on protected-area representation based on projected species distributions in 2070. We found that PAs added in the last three to four decades were better at representing species diversity than random additions overall. Threatened species, amphibians and reptiles are the exception. Species representation is projected to decrease across PAs under climate change, although PA expansions over the last decade (2006–2015) better represented species' future bioclimatic niches than did sites selected at random for most evaluated groups. These findings indicate an unbalanced representation across taxa, and raises concern over under-represented groups, including threatened species, and species’ representation under climate change scenarios. However, they also suggest that decisions related to locating protected areas have become more strategic in recent decades and illustrate that indicators tracking representativeness of networks are crucial in PA monitoring frameworks.