The success of data sharing among natural-history collections may alleviate the fears about capacity and cooperation expressed in your special issue (Nature 461, 145; 2009, Nature 461, 160–163; 2009 and Nature 461, 168–173; 2009). Our social and information-technology (IT) infrastructure provides open access to millions of records from hundreds of repositories, thanks to broad participation and funding from the US National Science Foundation. However, this success story also exposes some new challenges.

Key to this success has been the development of a distributed publishing system that conserves full rights of contributors to data and access. Data are curated at source and then made available to the community for use and improvement. The Mammal Networked Information System (MaNIS) was established in 2001 to provide hardware, data standards, transmission mechanisms and stakeholder support for open access to specimen data.

Since then, more thematic, taxonomic and regional networks have been created. Each of these feeds into a growing biodiversity-informatics community, including the Global Biodiversity Information Facility, which serves users with up to 189 million plant and animal records. The vertebrate-based biodiversity networks — including MaNIS, HerpNET, ORNIS and FishNet 2 — serve some 4% of their combined holdings each day to users hungry for these data.

Sustaining these resources is difficult. Growth has led to problems with scalability and sustainability, including difficulties in keeping resources running, slow provider response times and complicated installations and maintenance. The National Biological Information Infrastructure has provided support, and the vertebrate networks are consolidating into a platform called VertNet. In order to reduce IT costs, VertNet will move from institutional servers to a cloud computing platform, providing nearly unlimited room for growth.

But solving technological challenges is not enough. Our success has depended on strong engagement with our contributor and user community. Capacity building in biodiversity informatics is especially important. Success requires willing participation, robust technology choices and a commitment to engage fully with the communities these repositories will serve.