Historical data on biodiversity would be valuable for investigating the long-term impact of human activities. Contrary to popular belief, such data have been widely collected for several hundred years through initiatives that would today be described as 'citizen science'.

For example, in the late sixteenth century, the Spanish government distributed questionnaires known as relaciones topográficas to each village, with local inhabitants providing a compilation of natural-history knowledge. The 637 questionnaires that survive include information on some 190 species of wild animals and plants, gleaned from more than 4,300 individual records.

Also, the 16-volume geographical dictionary edited by the statistician and politician Pascual Madoz in the mid-nineteenth century contains information on most Spanish population centres, rivers and geographical landmarks. Madoz's enterprise involved the collaboration of more than 1,000 citizens and provided several thousand records of wild plants and animals, as yet unused.

Similar historical data sets also exist from China, most European countries and their former colonies. These can be biased towards socio-economically important species, but this bias can help to minimize misidentifications, for instance.

We call for an international, multidisciplinary effort (including historians, linguistic experts, geographers and biologists) to locate, compile and contextualize these invaluable historical data for incorporation into global biodiversity databases.