..A DIGITAL LIBRARY FOR AMAZONIAN SCIENCE

Nigel Pitman (ACA & ACCA), María del Carmen Loyola Azáldegui, Karina Salas (ACCA), Gabriela Vigo T. (Universidad Nacional Agraria La Molina) & Dave Lutz (Wake Forest University)

PROJECT DESCRIPTION

The scientific literature on Amazonian nature has a split personality. About half of what biologists have written is English-language text in peer-reviewed journals that are typically inaccessible to biologists in Amazonian countries; the other half is Spanish- or Portuguese-language text in books, theses, journal articles, and unpublished reports that rarely reach the developed world. Since half of the existing information on Amazonian biology is therefore invisible to any given researcher-even one who has access to state-of-the-art bibliographic resources-scientists working in the Amazon regularly repeat each other's experiments instead of building on them and talk past each other rather than exchanging ideas. In the face of accelerating development and deforestation across the Amazon basin, the last thing we need is bad science.

Figure 1. Growth in the number of scientific publications on the biology and conservation of Madre de Dios, Amazonian Peru, from 1971 to 2004 (N. Pitman et al., unpublished data).

In September 2004 we set out to address the problem on a small scale by compiling a list of everything ever written about the biology and conservation of Madre de Dios, a famously diverse Amazonian department in southeastern Peru. Six months later, thanks to a $3,500 grant from the Amazon Conservation Association, access to ISI Web of Science, and four Peruvian and American student assistants who rustled through >30 Peruvian libraries in three cities, we had assembled a bibliography of 2,050 books, articles, theses, and unpublished reports. Based on the declining returns of our searches by October 2005, we believe that the list now includes >90% of the texts ever written about the region: everything from 16th century explorers' accounts to papers published in peer-review journals a few months ago.

This is the first time that anyone has seen the entire body of research generated by 400 years' worth of observation by explorers, biologists, and conservationists in any part of the Amazon basin. Because 20% of the entries have full abstracts and all have keywords, it is now possible to locate all of the work on a certain species of animal, or from a certain national park, or by a certain researcher in Madre de Dios with a few keystrokes-something that is still beyond the reach of Google Scholar, the Library of Congress, ISI Web of Science or any other leading bibliographic resource. And because the list is databased in the popular reference management software Endnote (among other formats), our colleagues can put it to use the next time they write a paper about Amazonia.

Apart from its practical value for scientists, the bibliography is an interesting dataset in its own right. If the swelling flood of Amazonian publications (see Figure 1) is to form a cohesive body of work at the service of conservation, we need answers to questions like, How many texts on Amazonian biology include information of practical value for park managers? How many focus on abstruse topics with little relevance to conservation? Who is publishing the most-cited research? Who is funding that work? What proportion of the organisms studied by Amazonian biologists are globally threatened?

RESEARCH PRODUCTS TO DATE

In November 2005, thanks to John Janovec's Andes to Amazon Botany Project at the Botanical Research Institute of Texas, the bibliography was launched as a searchable, downloadable, web-based resource at

http://atrium.andesamazon.org/biblio_search.php

At the same address, visitors can download full copies of the database in .doc and .xml formats. The .xml file is easily imported into Endnote.

NEXT STEPS

Since the work to date has proven to be quick, cheap, and a good research opportunity for South American students, we are now aiming to broaden the project in five ways:

by extending the geographical reach of the bibliography to the rest of Amazonian Peru, as well as to Acre, Brazil, and Pando, Bolivia, with the help of Peruvian, Brazilian, Bolivian, and American students, colleagues, and institutions;
by extending the thematic reach of the bibliography to include key social texts, especially those on the history of indigenous communities, early European settlers, and the rubber boom (an excellent annotated bibliography of this kind was published for Madre de Dios in 1991);
by scanning and posting on the web especially important and hard-to-find texts, within the limits set by copyright law;
by ensuring that all entries in the database are fully indexed with trilingual kewords that will allow efficient searches in Spanish, English, and Portuguese, as well as developing connections to emerging web-based databases of plant and animal lists, photos, and researchers;
and by developing new tools and partnerships to make these resources accessible to the largest possible audience.